Left-wing populism on the rise?
An experimental study about the effects of populism in Dutch left-wing politics on persuasion and emotions
Jorien van der Weerd University of Amsterdam
Jorien van der Weerd 11353228
Master’s Thesis
Graduate School of Communication
Master’s programme Political communication science Linda Bos
28 January 2020 7493 words
Abstract
Populism is dramatically shifting the global spectrum. This change influences mainstream parties, as they adopt a populist style and rhetoric. But is this effective? Most studies focus on the effects of right-wing populism, but little is known about left-wing populism. Therefore, this study focusses on the effects of the use of a populist style in left-wing politics. The aim of this study is to investigate the effects of a populist message from a left-wing populist party and a left-wing mainstream party on persuasion and evoked emotions.
This study has a 2x2 experimental design (N = 167) and manipulates the message (populist versus neutral) and the source (populist party versus mainstream party) in a Twitter post. This study contributes to the existing literature in two ways. First, it shows that the adaptation of a populist style by a mainstream left-wing party creates a backlash effect. Secondly, the use of a populist style by the populist party also creates an adverse reaction. Therefore, using a neutral style and tone is obviously more credible and increases voting behaviour for the populist and mainstream left-wing parties in the Netherlands.
Keywords: populism, political communication, experimental research, left-wing populism, persuasion, emotions
Left-wing populism on the rise?
Last October 2019, the UN secretary-general António Guterres gave an outstanding speech in which he mentioned the remarkable increase of demonstrations around the world. Some protests were related to economic issues, others for political demands, human rights, discrimination and climate change. But in essence, all of the above can be connected to the growing deficit of trust between people and political establishments, and rising threats to the social contract. People face rising inequality and want their human rights respected, but above all, they want to be heard (Guterres, 2019). Interestingly, populist leaders are precisely
tapping into these problems and peoples’ dissatisfactions. By doing so, they try to gain trust and make people believe that they are the chosen ones who will create the change people are asking for.
The growing influence of populist parties is increasingly recognizable around the globe, populist leaders (e.g. Donald Trump, Boris Johnson, Victor Orbán, Jair Bolsonaro) take power on the political stage. Mudde (2007) argues that the success of populist parties has changed the political spectrum. He argues that the populist discourse has become visible in mainstream politics of Western democracies. Mudde (2013) therefore also wonders if populism is contagious, as populist themes and rhetoric are adopted by mainstream parties coined ‘soft populism’.
In the Netherlands populist parties like Forum voor Democratie (FvD), Partij voor de Vrijheid (PVV) and the Socialistische Partij (SP) are successful in elections. Academic literature has investigated how populist parties, populist politicians and/or leaders spread their messages (Hameleers & Schmuck, 2017; Kramer, 2017; de Vreese et al., 2018). Therefore, we know that populist communication can be persuasive and evokes negative emotions among its receivers (Aalberg et al. 2017; Bos, van der Brug & de Vreese, 2013; Jagers & Walgrave, 2007; Mudde, 2007). Strong emotions and feelings of fear and risks can have an
important influence on the formation of attitudes, opinions, political preferences and also voting behaviour among individuals (Schmuck & Matthes, 2017; Wirz et al., 2018).
While we know much about the effects of populist communication by populist politicians, we do not know if mainstream parties could also profit by using this type of communication. Considering the growing influence of populism, the question rises if mainstream parties could also benefit from using a populist style just like populist parties? We do know that populism can be persuasive, but can mainstream parties also use ‘soft populism’ strategically? Can mainstream parties incorporate populist communication in their political marketing to gain votes or win elections?
When looking at the scientific relevance, most literature is focused on right-wing populism. Studies investigating left-wing populism in the Netherlands are lacking. The study of Bos and colleagues (2011) investigated if the effects of using a populist style and rhetoric by a mainstream right-wing party was effective. To advance the field, this study will examine the same but will focus on the left side of the political spectrum. Therefore, the following research question is proposed:
“What are the effects of a populist message from a populist party and a mainstream party on persuasion and evoked emotions?”
To answer the research question, an online survey embedded experiment will be conducted with a 2x2 factorial design.
Theoretical Framework Populism as ideology
The increasing visibility and growing popularity of populist leaders over the world leads to more curiosity about populism and what it exactly entails. Scholars describe populism as an ideology using a variety of definitions (Abou-Chadi, 2016; Canovan, 2002; Jagers &
Walgrave, 2007; Mudde, 2000). Mudde (2007) describes the key features of populism as an ideology which separates society into two homogenous and antagonistic groups, namely, ‘the pure people’ versus ‘the corrupt elite’ and argues that politics needs to be an expression of the general will of the people. Populists blame the elite for not listening and acting upon the will of the people, and critiques the role of the elite only pursuing their own interests (Mudde, 2007, p.23). Taggart (2006) talks about the ‘heartland’ within the populist ideology. The heartland presents the ideal world which is constructed from the past and portrayed as lost in the current situation. In essence, the heartland is the good life of the people that was there in the past, but is gone due to corruption and social changes in the present. The elites are blamed to be responsible for this (Taggart, 2006).
The heuristic model of populist ideology by Engesser, Fawzi and Larsson (2017) shows the interrelations of the most important elements of a populist ideology, namely, the populist, the elite, the people and ‘others’ (Figure 1). This model shows the fight for political
dominance, where the people are the ones who need to own the right of political dominance. The populist is seen as the challenger of the elite and as a defender of the rights of the people. The people play a crucial role in the ideology of populism. They are seen as a homogenous group, the ingroup (Reinemann et al., 2017, p.19). The populist aims to take away the power of the elite and give it back to the people. The last element, ‘the others’, engage with the elite and are seen as a rival of the people. They are the anti-outgroup. The populist wants to exclude ‘the others’ from society. The outgroup are usually immigrants in right-wing populism. In left-wing populism, the key outgroup are mostly the capitalists, the super-rich (Engesser et al., 2017).
Right-wing and Left-wing populism
Most literature is focused on right-wing populism, yet a significantly difference can be discerned between right and left-wing populism (Mudde & Taggart, 2019). Right-wing
populism is mostly focussed on anti-immigration, nationalism, and the idealization of the heartland. The outgroup within right-wing populism are mostly immigrants who disturb the social harmony (Mudde, 2007; Taggart, 2000). Left-wing populism sees the superrich and capitalists as the outgroup, which are a threat to the people. Left-wing populism blames the rich and capitalists for being self-serving, corrupt and egoistic. The political elite is blamed for taking away the resources of the people and are accused of prioritizing the needs of the outgroup (capitalists) above the ordinary, hard-working good citizens. Besides blaming the rich, a great emphasis lies on the already negatively perceived and socially less accepted outgroups within the left-wing populist setting (Hameleers et al., 2018). So within populism, there are two different directions to consider, left and right. Both see different outgroups as a threat and hold different outgroups accountable for the injustice faced by the ingroup
(Hameleers et al., 2018; Mols, 2012). Populist political parties
When looking at populist parties, we see that they act differently than mainstream political parties. Taggart (2006) points out that populist parties argue that the ‘pure people’ suffer from the leadership of the elite. Populist parties’ rebel against the current politics and take upon a political colour like a chameleon based on the desires of the people at a certain time. According to the model of issue competition, also named saliency model, political parties stress issues which fit within their core values and avoid issues of competitors (Budge & Farlie, 1983; Stier et al., 2017). Populist parties mostly lack these core values and fill this gap by using powerful statements which focus on what needs to change within society. Moreover, populist parties are called the ‘people’s parties’ as they say what the people want
to hear. It is mostly in times of crisis that populist parties respond to the interest of the people and consider themselves as crisis managers (Taggart, 2006).
When looking at the political environment in the Netherlands, Rooduijn and colleagues (2019) identify three populist parties. First of all, the right-wing populist party PVV (Partij van de Vrijheid). This party is against the Islamization of the Dutch society, has an anti-European sentiment and wants to give the power back to the people (PVV, n.d.). Secondly, the left-wing populist party SP (Socialistische Partij). This socialist party strives for justice, equivalence and solidarity for everybody. This party strives for equality in society (health care, education, income, employment) and is against capitalism (SP, n.d.-a). The last populist party in the Netherlands is FvD (Forum voor Democractie). The party is established in 2016 and is still upcoming. The party has an anti-elite sentiment and their main standpoints are focused on nationalism, anti-elites, immigration and anti-EU. They have a strong focus against the complete political system (FvD, n.d.).
Populist communication
Experimental studies have shown that populist communication can activate or prime citizens’ political perceptions (Hameleers et al., 2017). Albertazzi and McDonnell (2008) define the communication style of populists as the ‘man in the street communication style’. But also as a communication style that is clear and straightforward by using simple and strong language. Populist communication can be recognized by dramatized messages which generate tension. Furthermore, populist communication is known by fast and intense language and can be seen as ‘friend versus foe’ rhetoric (Albertazzi, 2007; Schedler, 1996; Taggart, 2000; Weyland, 2001).
In addition, the populist communication style includes personalization,
emotionalization and anti-establishment attitudes (Ernst et al., 2018). According to Taggart (2000), the focus of a populist communication style lies on the anti-political establishment,
anti-elitism and a celebration of the ‘nation’, the ‘heartland’ and the ingroup. To illustrate, populists spread messages which stress that the ingroup is threatened. They do this by
communicating that the elites, others or social outgroups are taking away what belongs to the people and what they deserve (Elchardus & Spryt, 2016; Hameleers et al., 2018; van Spanje, & de Vreese, 2015). Furthermore, populist parties blame mainstream parties by arguing that they do not connect and listen to ‘the people’ (Bos et al., 2012). Populism highlights a strong sense of urgency of the threats faced by ‘the people’ and spread messages about failing of the elites and the dangers of the social outgroups.
Persuasion
Experimental studies have shown that populist communication can be very persuasive and can activate and prime citizens political perceptions (Hameleers et al., 2017; Rooduijn et al., 2014; Schmuck & Matthes, 2017). The nature of persuasion is the intention to change attitudes and influence people’s minds (Kwak et al., 2018; Rocklage, Rucker & Nordgren, 2018; Thorson, 2014). Persuasion has been considered as an important aspect of political communication by influencing public opinion, political behaviour and voting behaviour (Dowding, 2016; Kwak et al., 2018). Thorson (2018) found that persuasive communication can lead to creating awareness on opposing political positions and clarifies the candidates’ positions on issues. Moreover, populist communication explicitly emphasizes that the
ingroup, ‘the people’, are threatened which can lead to mobilization, political engagement and thus persuasion of messages.
When looking at left- and right-wing populism, different scapegoats are held accountable for the injustice and threats the ingroup is facing. Blame shifting of societal problems plays a central role within populist communication, where the ingroup of ordinary citizens suffer from societal problems caused by the outgroup. Populist messages that blame
political elites or outgroup for societal issues are more likely to be persuaded by the ingroup than when these messages do not blame certain outgroups (Bos et al., 2019).
Hameleers and Schmuck (2017) show that messages blaming political elites or outgroups are more persuasive among recipients who identify with the source of the message. The ingroup or reference group is thus very important. In addition, ingroup cues in
communication have shown to increase the credibility and persuasiveness of messages by ingroup sources. Therefore, ingroup recipients are more likely to agree with the issue position the populist message suggests (Hogg & Reid 2006). Considering the literature on populism, a populist communication style and the empirical findings of the relationship between populist messages and persuasion, the following hypothesis is proposed:
H1: Exposure to a populist message will be more persuasive than exposure to a neutral message.
Emotions
Most populist parties use extreme negative slogans and images that stigmatize
outgroups. These slogans are accompanied by fear elicit visual images with stereotypical and negative presentations of the outgroup which evoke emotions among its receivers, ‘the people’ (Schmuck & Matthes, 2017). Populist parties use these images to trigger threat appeals of the ingroup such as their social values, traditions, economic and social welfare (Atwell Seate & Mastro, 2015). Emotions are thus an important factor within populist communication. The use of certain frames within populist communication can elicit positive emotions, like hope, strength and advocacy, or negative emotions, like fear and anger (Lecheler, Schuck, & de Vreese, 2013). Moreover, the evoked emotions through populist communication are very effective in influencing the formation of attitudes and opinions of individuals (Rocklage, & Fazio, 2016;Schmuck & Matthes, 2017; Wirz et al., 2018). To illustrate, experimental evidence suggest that the activation of positive emotions can lead to pride and social responsible behaviour. In contrast, negative emotions can lead to feelings of
anxiety and anger which can lead to mobilization, a negative evaluation of outgroups and polarization in society (Wirz, et al., 2018). The activation of emotions via populist
communication can thus have powerful influence on individuals political attitudes and voting behaviour.
Furthermore, the use of images can highlight specific relevant information to the receiver and can elicit emotions in processing the information. So a populist frame presented in image and text will lead to a very strong effect in evoking emotions (Rocklage &
Nordgren, 2017). To illustrate, a message which negatively frames immigrants as a threat via image and text, so with a strong image-text congruency, can result in more negative attitudes towards immigrants (Schmuck & Matthes, 2017). Based on the perception that populist communication uses strong language and textual images to evoke emotions, the following hypothesis is proposed:
H2: Exposure to a populist message will evoke stronger emotions than exposure to a neutral message.
Issue ownership
Political messages can have a considerable impact on the formation of opinions about certain issues. Populist parties use political message to inform the public about candidates, political issues and put emphasis on the threats the ingroup is facing (Bale et al., 2010; Schmitt-Beck & Farrel, 2002). Not only populist parties but also mainstream political parties stress the importance on issues, as they strategically communicate about certain issues to persuade voters (Bos, Lefevere, Thijssen, & Sheets, 2017; Holian, 2004). This is called issue ownership. Issue ownership theory argues that an issue is owned by a party. When parties or political candidates get media attention, they will be better known by the specific issue they stand for, and this will increase their party visibility. Which in turn can change voters’ behaviour so they will vote for the party. The multidimensional concept of issue ownership
consists of a ‘competence’ dimension and an ‘associative’ dimension. The competence dimension of issue ownership focusses on the perception that certain parties can handle a certain problem better and have the ability to solve or deal with an issue. The associative dimension focusses on the associations that parties have with a certain issue. Which in turn leads that easily retrievable information comes first on top of voters’ minds in the
information-processing of the brain (Scheufele and Tweksbury 2007). Associative issue ownership is thus about the connection of an issue to a party, and can therefore have cognitive effect on people’s voting behaviour when certain issues are salient (Walgrave, Lefevere & Tresch, 2012).
To illustrate, the issue on economic equality is a very important issue for populist left-wing party SP. They argue that the rich need to pay more taxes in order to create more equality in society. Thefocus on this topic is explicitly connected to this party. They have issue ownership on this issue in a sense that when people hear about this issue, they directly connect this to the SP (associative issue ownership). When you ask which party could solve the issue the best, they will also refer to the SP (competence issue ownership).
Literature about issue ownership suggests that parties will emphasise the issue they own and will ignore issues by other parties (Bos, et al, 2017; Budge, 2015). Owning certain issues gives parties reputational advantages as the media can have a clear focus on the issues which are connected to the party. The study of Bos and colleges (2017) examined the
relationship between exposure to media on various issues and citizens’ support for a populist party. The results show that when a party is mentioned in connection to the issue they own, this leads to a positive effect on the party support.
In addition, populism can also be seen as an issue, as it can be directly linked to a certain party. An experimental study shows that populism is owned by the Dutch right-wing populist party PVV. When the right-wing mainstream party VVD adopts a populist style and
rhetoric, this leads to a negative evaluation. Whereas the populist party PVV was not negatively affected for using a populist style and rhetoric (Bos, et al., 2011).
Furthermore, Sheets, Bos and Boomgaarden (2015) argue that awareness of the potential power of populist discourses on all voters is needed. Therefore, in this thesis I argue that populism itself can be seen as an issue owned by populist parties, at least more than mainstream parties. To test this assumption the following hypotheses are proposed:
H3a: Exposure to a populist message from a populist party will be more persuasive than exposure to a neutral message.
H3b: Exposure to a populist message will be more persuasive than exposure to a neutral message, but this effect will be stronger for messages from a populist party than messages from a mainstream party.
Considering the effects of a populist message on evoked emotions, can the same be assumed as for the hypotheses on persuasion? Existing literature did not give enough insights for the relationship between issue ownership and emotions. By hoping to expand insights in this area two research questions are proposed:
RQ2a: Will exposure to a populist message from a populist party evoke stronger emotions than exposure to a populist message from a mainstream party?
RQ2b: Will exposure to a populist message evoke stronger emotions than exposure to a neutral message, and will this effect be stronger for messages from a populist party than messages a from a mainstream party?
Method
Research design
To test the hypotheses and answer the research question an online survey embedded experiment was conducted. The experiment had a 2x2 factorial design. The 2 (message: populist and neutral) x 2 (source: populist party and mainstream party) design allowed for comparing the effects of type of message and type of source. An experiment provides the possibility to manipulate the content of the message and allows to conduct a causal test. The use of an experimental design offers the opportunity to test the direct effects of the specific media content. Furthermore, it allows to test the direct effects on emotions and make the comparison between the experimental group and the control group. Via this research design an answer can be given on how the public reacts if a mainstream political party uses a populist style. Also, former studies who investigated the effects of populist messages on citizens perceptions, attitudes and/or emotions used an experimental design (Blassnig & Wirz, 2019; Bos, de Vreese, & van der Brug, 2013; Hameleers & Schmuck, 2017).
Sample
The participants (N = 167) in the study had to be at least 18 years old in regards to the General Data Protection Regulation (GDPR). Furthermore, the participants had to live in the Netherlands, in order to understand the political situation in the country and be able tovote. The recruitment of the participants was based on a non-probability sample. The respondents were recruited via the personal network of the researcher by the use of online communication channels, such as Facebook, Whatsapp and LinkedIn. The study aimed for a well distributed sample that represents the Dutch population regarding gender, age and education.
< Table 3. about here > < Figure 2. about here >
Procedure
The online experimental survey started by informing the participants about ethics and insurance of their anonymity. The respondents had to give informed consent as a requirement for participating in the study. The informed consent required that participants were willing to participate and had an age of 18 years or older. The ethical approval of the study was
provided by ASCoR of the University of Amsterdam. The survey was administered in Dutch and the data collection lasted from 29 November until 11 December 2019. To make sure that the manipulations were perceived as how they were intended, a pre-test was conducted. Thereafter, the data collection started (information about the pre-test can be found in Table 1 and Table 2 in Appendix I). After accepting the informed consent, questions regarding participants’ demographics and political interests were asked. The survey then proceeded to the exposure of the stimulus material. The participants were asked to take a look at the political message for at least 10 seconds. Followed by a series of questions regarding evoked emotions as a result of seeing the stimuli. Thereafter, respondents were exposed to political statements to measure the level of persuasion, accompanied by a question that asked the probability participants would vote for the party they saw in the message. The last part of the survey consisted of a manipulation check in order to see if the participants perceived the stimulus material in the right way. At the end of the survey the participants were thanked for their participation and debriefed about the factiousness of the messages and the purpose of the study.
Stimuli
Each group was presented one out of four Twitter posts that were designed for the purpose of this study. The Twitter post contained a message addressing the problems within the education system in the Netherlands. Each participant was randomly assigned to one of the four experimental conditions:
(1) populist message from populist party SP (2) neutral message from populist party SP
(3) populist message from mainstream party GroenLinks (4) neutral message from mainstream party GroenLinks.
The populist message (experimental group) blamed the government for the problems within the education system and had a strong populist style and rhetoric. The failing of the government, blame on the capitalists and failing of the system was explicitly mentioned. The neutral message (control group) only addressed the problems within the education system without explicitly blaming the government for the problems. The populist style and rhetoric was absent in this message. The political messages were based on already existing texts about education from the websites of the political parties (GroenLinks, n.d.; SP, n.d.-b)
Part of the populist message:
“Shortage of teachers disastrous for future of our children!
It is ridiculous that children are send home due to the shortage of teachers…It is unbelievable that the growing economy delivers the government millions but they abandon our teachers. The capitalists are their number one priority and politics only thinks about themselves. What are they doing?... We have to invest in our children, in our future!”
Part of the neutral message:
“Shortage of teachers stays a problem
Children face the problem of dropped classes due to a shortage of teachers… The government earns enough money due to the growing economy. More needs to be invested in the education sector to solve these problems and provide good education…Good education is an investment in the future.”
The message participants were exposed to could be from the populist party SP or the
mainstream party Groenlinks. The messages were created as a tweet from the twitter accounts of the political parties. Politicians and political parties are often associated with the successful use of social networking sites (Blassnig & Wirz, 2019). Twitter can be considered as a
medium that allows for an effective virtual political space where information, and politics are freely discussed, debated and exchanged. This platform is highly used by politicians and political parties (Hong & Nadler, 2011). The four types of stimuli can be found in Appendix II.
Data cleaning
205 people agreed to participate in the study. Of these 205 participants, four had to be excluded from the study because they indicated not meeting the criteria of being older than 18 years or having the right to vote in the Netherlands. Another 34 participants were removed from the final dataset due to missing values and not finishing the survey. Analyses were conducted over a final sample of 167 participants.
Randomization check
A randomization check showed that the randomization worked. The conditions did not differ with regard to Age F(3,167) = 0.27, p = .845, η2 = 0.05, Gender F(3,67) = 0.57, p = .639, η2 = 0.01 and Education level F(3,167 ) = 1.41 , p = .241 , η2 = 0.03. Furthermore, the manipulation check showed that the four conditions did not differ across Political interest F(3,167) = 0.82, p = .487, η2 = 0.02 and Political identification F(3,167) = 1.02, p =, .387, η2 = 0.02.
Manipulation check
The credibility of the stimulus material was good with a high mean score and the conditions did not differ much in their level of credibility (Table 4). This was checked via the
question: ‘To what extent would you expect this message on the social media channels of the political party?’on a 7-point scale (1 = totally not, until 7= totally).
< Table 4. about here >
In order to make sure that the experimental conditions were perceived as how they were intended to be perceived, some control questions were asked (Table 5). The
manipulation check was measured with statements regarding the populist tone of the message on a 7-point scale (1= totally not applicable, until 7 = totally applicable).
< Table 5. about here >
To check if participants recognized the party in the message, the following question was asked: ‘Was the message you just saw from the following party?’ on a 7-point scale (1= totally not applicable, until 7 = totally applicable). The manipulation check for type of party worked (Table 6). Participants recognized the party they saw in the message.
< Table 6. about here > Variables
Conditions. The experimental conditions were measured by using four different types of messages. These experimental conditions were first recoded into, 1 = SP populist (N = 43), 2 = SP neutral (N = 43), 3 = GL populist (N = 39), 4 = GL neutral (N = 42), thereafter they were computed into the one variable named Conditions. Furthermore, the conditions were recoded into the variables Parties (1 = SP, 2 = GroenLinks) and Message (1 = populist message, 2 = neutral message). For the regression analysis the variables were recoded into Type of party (0 = GroenLinks, 1 = SP) and Type of message (0 = neutral, 1 = populist).
Persuasiveness. The persuasiveness of the message was assessed via the two dependent variables Level of persuasion and Probability to vote. Level of persuasion was measured by six statements on a 7-point scale. The items tapped how much participants (dis)agreed with the populist statements. The items were (1)‘More money needs to go to
education, even when this leads to less budget for other sectors’, (2)‘The government has already invested enough money in the education sector. It is now to the education sector itself’, (3)‘The teachers have to strike until there are structural changes in the education sector’, (4)‘The government fails to address the problems in the education sector’, (5)‘Due to the teachers’ strikes, there are only more problems, and lessons are cut’ and (6)‘The
government is not responsible for the problems in the education sector’.
Because all statements were not formulated in the same direction, statement 2, 5 and 6 were first reverse recoded. Thereafter, an Exploratory factor analysis was conducted
(principal axis factor analysis with Varimax). Both the Eigenvalue-criterion and Screeplot showed that there are two factors (eigenvalue factor 1: 2.39; eigenvalue factor 2: 1.09). Together they explained 58.13% of the variance. The factor analysis showed a weak reliability of the scale with a Cronbach’s alpha of 0.68. If statement 5 was deleted, the Cronbach’s Alpha was 0.70. Therefore, statement 5 was deleted. An Exploratory factor
analysis and Screeplot with the five items measuring (statements 1,2,3,4,6) level of persuasion showed only 1 factor (eigenvalue factor 1: 2.32), which explained 46.31% of the variance and had a good reliability with a Cronbach’s Alpha of 0.70. The five items were computed into a scale measuring Level of persuasion. Furthermore, the second dependent variable Probability to vote was measured by asking the probability a participant would vote for the party
presented in the experimental condition on a 7-point scale (1= not likely at all, until 7= very likely).
Emotions. The dependent variable Emotions was measured via the two variables Feeling and Evoked emotions. Feeling was measured by a question asking respondents to indicate the level of feeling positive/negative after seeing the stimuli on a 7-point scale (1= negative, 4 = neutral, until 7= positive).
Furthermore, the variable Evoked emotions was measured by indicating the level of evoked emotions after seeing the stimuli for the emotions: anger, frustration, hope, sadness, no emotion and other emotions on a 7-point scale (1 = describes my feeling totally not, until 7 = describes my feeling completely). 1
Results
In order to check what the effects of a populist message from a populist party and a mainstream party are on persuasiveness and emotions, the following hypotheses are tested. The first hypothesis predicted that exposure to a populist message will be more persuasive than a neutral message. To test this assumption an independent samples t-test was conducted with Type of message (populist versus neutral) as independent variable and Level of
persuasion and Probability to vote as the dependent variables.
Level of persuasion. An independent samples t-test was conducted to examine whether a populist message scored higher on the level of persuasion than a neutral message. Levene’s test for equality of variance was not significant, F = 0.83, p = .364. The independent samples t-test indicated that the level of persuasion after seeing the message in the populist conditions (M = 5.02, SD = 1.05) did not differ significantly from the neutral conditions (M = 4.97, SD = 0.98), t = 0.29, p = .772.
Probability to vote. When looking at the probability participants would vote for the party they saw in the message, and independent samples t-test was conducted. Levene’s test for equality of variance was not significant F = 0.76, p = .386. A significant difference between the populist conditions (M = 2.83, SD = 1.55) and the neutral conditions (M = 3.74,
1Other emotions indicated by participants were: regret, worry, irritation, powerlessness, incomprehension, impatience, unbelief, uncertainty, combativeness, disappointment and surprise.
SD = 1.80) was found, t = -3.46, p = .001. Participants who saw the neutral message were more likely to vote for the party in the message than participants who saw the populist message.The results for H1 indicate that the opposite of what was predicted was found, therefore H1 is rejected.
The second hypothesis predicted that exposure to a populist message will evoke stronger emotions than exposure to a neutral message. To test this assumption an independent samples t-test was conducted with Type of message (populist versus neutral) as independent variable and Feeling and Evoked emotions as dependent variables.
Feeling. An independent samples t-test was conducted to examine whether a populist message scored higher on the feeling positive/negative than a neutral message. Levene’s test for equality of variance was not significant F = 0.63, p = .429. The independent samples t-test indicated that the level of feeling positive/negative after seeing populist messages (M = 3.54, SD = 1.21) did significantly differ from the neutral messages (M = 4.02, SD = 1.46), t = -2.38, p = .021. This means that respondents who saw the populist messages felt more negative in comparison with respondents who saw the neutral messages.
Evoked emotions. An independent samples t-test was conducted to examine whether populist messages evokes stronger emotions than a neutral messages. Results are shown in the Table 7.
< Table 7. about here >
The independent samples t-test did not show a significant difference between exposure to populist messages and a neutral messages on the level of evoked emotions. Only a
significant negative effect was found for feeling positive/negative after exposure to the two different stimuli. Some support is found for H2.
Hypothesis 3a predicted that exposure to a populist message from a populist party will be more persuasive than a neutral message from a populist party. To test this assumption an independent samples t-test was conducted with Level of persuasion and Probability to vote as dependent variables and SP message (populist vs neutral) as independent variable.
Level of persuasion. Levene’s test for equal variance assumed showed to be significant F = 4.32, p = .041. The independent samples t-test indicated that level of persuasion after seeing the populist message (M = 5.04, SD = 1.18) did not significantly differ from seeing the neutral message (M = 4.97, SD = .90) from the SP, t = 0.27, p = .787.
Probability to vote. Looking at the probability to vote showed the Levene’s test for equal variances assumed not to be significant F = 0.10, p = .752. The independent samples t-test showed that there is a significant difference in the probability to vote after seeing the populist message (M = 2.75, SD = 1.55) and the neutral message (M = 3.74, SD = 1.79) from the SP, t = -2.70, p = .008. This indicates that participants who saw the neutral message from the SP were more likely to vote for the party than participants who saw the populist message from the SP. The results indicated that no support was found for H3a.
Hypothesis 3b predicted that there is an interaction effect. Exposure to a populist message will be more persuasive than a neutral message, but this effect will be stronger for messages from a populist party than from a mainstream party. To test this assumption a regression analysis was conducted with Level of persuasion and Probability to vote as
dependent variables and Type of message (0 = neutral message and 1 = populist message) and Type of party (0 = GroenLinks and 1 = SP) as independent variables.
Level of persuasion. A regression model with level of persuasion as dependent variable and type of party and type of message as independent variables was not significant, F(3,165) = 0.04, p = .991. Type of message had no significant association with the level of persuasion, b = 0.03, t = 1.25, p = .901. Type of party had no significant association with the
level of persuasion b = 0.01, t = 0.03, p = .998.The interaction effect of type of party*type of message on the level of persuasion had no significant association b = 0.04, p = .998. The regression model explained 2,6% of the variance in Level of persuasion (R2 = .026).
Probability to vote. A regression model with probability to vote as dependent variable and type of party and type of message as independent variable was significant, F(3,165) = 3.96, p = .009. Type of message had a negative significant association with the probability to vote, b = -0.88, t = -3.15, p = .033. Type of party had no significant association on the probability to vote b = 0.06, t = 0.02, p = .987. The interaction effect of type of party*type of message did not show to have a significant association with the probability to vote b = -0.18, t = -0.33, p = .739. The model explains 26,4% of the variance in probability to vote (R2 = .264). These results indicate that no support for H3b was found.
Research question 2aasked if exposure to a populist message from a populist party will evoke stronger emotions than a populist message from a mainstream party. To test this assumption an independent samples t-test with Feeling and Evoked emotions as dependent variables and Populist SP and Populist GroenLinks as independent variables was conducted.
Feeling. Levene’s test for equal variance assumed showed not to be significant F = 0.05, p = .823. The independent samples t-test indicated that feeling negative/positive after seeing the populist message from SP (M = 3.49, SD = 1.21) did not significantly differ from seeing the populist message from GroenLinks (M = 3.59 , SD = 1.23), t = -0.37, p = .709.
Evoked emotions. An independent samples t-test was conducted to examine whether a populist message from SP evoke stronger emotions than a populist message from GroenLinks. Results are shown in the Table 8.
< Table 8. about here >
The independent samples t-test did not show a significant difference between exposure to the populist message from SP or the populist message from GroenLinks on the level of
evoked emotions. To answer the research question, exposure to a populist message from a populist party does not significantly evokes stronger emotions than a populist message from a mainstream party.
Research question 2b asked if exposure to a populist message evokes stronger emotions than a neutral message, and if this effect will be stronger for messages from a populist party than from a mainstream party. To test this research question a regression analysis with Feeling and Evoked emotions as dependent variables and Type of message (0 = neutral, 1 = populist) and Type of party (0 = GroenLinks, 1 = SP) as independent variables was conducted.
Feeling. A regression model with feeling as dependent variable and type of party and type of message as independent variable was significant F(3,165) = 3.13 p = .027. Type of message had a significant negative association with feeling positive/negative b = 0.72, t = -2.43, p = .016. Type of party had a negative significant association with feeling
positive/negative b = -0.57, t = -1.95, p = .053. The interaction effect of type of party*type of message did not show to have a significant association with the variable feeling b = 0.46, t = 1.11, p = .267. The model explains 23,5% variance in the variable feeling (R2 = .235).
Evoked emotions. Regression models with the types of emotions as dependent
variables and type of party and type of message as independent variables are shown in Table 9.
< Table 9. about here >
The regression analysis did not show any significant differences for the interaction effects of type of party*type of message on level of evoked emotions. Both regression analyses did not show a significant effect for RQ2b. The effect of a populist message on feeling and emotions is not moderated by partisan source.
Conclusion & discussion
The overall purpose of this study was to investigate the effects of populist messages from a left-wing populist party and a left-wing mainstream party on persuasion and evoked emotions via an online survey experiment. So far, little was known about the effects of left-wing populism on persuasion and emotions. The results indicated that use of a populist style does not persuade. On the contrary, exposure to a populist message decreased the probability to vote and led to more negative feelings. Voters were not more persuaded by a populist message nor were they more emotionally aroused. Moreover, the source of the message did not moderate these effects.
The first objective of the study was to investigate the persuasiveness of populist messages. The findings of the study showed the opposite effect, namely that the use of a populist style has negative effects on the persuasion of the message. These findings are in line with the study of Hameleers and Schmuck (2017), who argue that the adaptation of a populist style can elicit a ‘backlash’ effect on recipients who do not support populist ideas. This may be one of the reasons why this study found this results, as it seems that the adaptation of a populist style created an adverse reaction by the recipients.
The second objective of the study was to look at the effects on emotions in terms of feeling positive or negative after seeing a populist message and evoked emotions like anger, fear, hope and sadness. The results show that recipients felt more negative after seeing the populist message. This does echo previous studies who argue that populist communication increases and triggers mostly negative feelings and emotions (Lecheler, Schuck & de Vreese, 2013; Schmuck & Matthes, 2017; Wirz et al., 2018). Although no significant effects were found for exposure to a populist message on discrete emotions, the mean scores showed that exposure to a neutral message leads to higher levels of evoked emotions, whether negative or positive. This might be explained by the fact that only five types of emotions were measured.
Hereby, recipients could only indicate their emotion to these five types, which they may not have felt or found difficult to indicate on a seven-point scale. Also, other types of emotions were filled in by the respondents like; regret, worry, irritation, powerlessness,
incomprehension, impatience, unbelief, uncertainty, combativeness, disappointment and surprise. Measuring emotions differently thus might have led to different results.
When looking at the effects of a populist message and the source (populist party versus mainstream party) a negative effect was found. When the populist party SP used a populist style, recipients perceived this as less credible than a neutral style. This is surprising as previous research indicates that the use of a populist style and rhetoric does not harm but also does not have a positive effect on a right-wing populist party (Bos et al., 2011). The present study found a different result when looking at left-wing populism. When the left-wing populist party SP used a populist style, a significantly negative effect was found on
probability to vote. This means that the party is more negative affected when using a populist style than a neutral style.
Moreover, this study and the study of Bos and colleges (2011) show similarities in findings. Both studies found that mainstream parties are dismissed when using a populist style. However, a different result was found in the present study, the use of a populist style by the populist left-wing party also created a reverse reaction. The findings of this study indicate that the adaptation of a populist style by left-wing mainstream and left-wing populist parties does not seem to be a good strategical action as it leads to a backlash effect.
These results lead to the question if left-wing populism has a stronger backlash effect in comparison with right-wing populism. Or do people criticise left-wing parties differently than right-wing parties? Another argument for this result could be that respondents not evaluate the left-wing populist party SP as populist, at least not as populist as right-wing
populist parties (Kessel, 2011). Future research should investigate the differences between left and right-wing populism and the effects of adaptation by mainstream parties.
This study is not without limitations. First of all, recipients were only shortly exposed to the message. Long-term repeated exposure to the messages could lead to different results (Lecheler, Schuck & de Vreese, 2013). Secondly, the topic of this experimental survey was related to the problems in the Dutch education sector. At the time of data collection, this topic received a lot of media attention which may have led to external influences. Therefore, it is recommended for future research to use a less salient issue as recipients will be less familiar with the topic, this can lead to different results.
Third, most people in the sample were on average highly educated, moderately politically interested and political centre-middle, right-wing oriented. Only 2% voted for the populist party SP and 18.8% for the mainstream party GroenLinks. Future research should use a more diverse and representative sample of the population. By the use of stratified sample the political interest of society will be correctly represented by the sample. It would be interesting to investigate which type of respondents react positive or negative to populist messages, in order to get a better identification of the type of individuals who are persuaded by populist messages.
To conclude, this study demonstrates that a populist message used by left-wing parties leads to negative feelings and a lower probability to vote by recipients. The study builds on previous studies, that support the idea that populist communication will only be accepted when citizens identify with the party that communicates the message (Bos et al., 2011,
Hameleers & Schmuck, 2017; Matland & Murray, 2013). When a mainstream left-wing party adopts a populist style, a backlash effect emerges. This is line with literature on ingroup serving bias and indicates that perceived similarity with the source is a crucial indicator for acceptance of the message (Householder & LaMarre, 2004). Moreover, this study shows that
populism cannot just be easily adopted by left-wing mainstream parties. Respondents reacted more negative to the populist message when send by a left-wing mainstream party, but also when send by a left-wing populist party. This indicates incorporating a neutral style in the political marketing of the populist party SP will be more effective in terms of voting
behaviour. Furthermore, the findings of the study are a contribution to the knowledge about populism and can be seen as a positive finding. As mentioned earlier in the study, Guterres (2018) argues we have to watch out for more polarisation in our societies. This study gives a positive note that incorporating populism in political marketing strategies will not increase party credibility or gain votes. The populist zeitgeist of Mudde (2013) and the adaptation of soft populism is not found to be effective in left-wing politics. Strong mainstream parties who connect to the people and have the ability work to on factors that create a growing deficit of trust instead of creating hate and anger, can save our democracies from threats to the social contract.
References
Abou-Chadi, T. (2016). Niche party success and mainstream party policy shifts - How green and radical right parties differ in their impact. British Journal of Political Science 46(2), 417-436. doi:10.1017/S0007123414000155
Albertazzi, D. (2007). Addressing ‘the people’: a comparative study of the Lega Nords’ and Lega Dei Ticinesi’s political rhetoric and styles of propaganda. Modern Italy 12(3), 327-347. doi:10.1080/13532940701633791
Albertazzi, D., & McDonnell, D. (2008). Twenty-first century populism: The spectrum of Western European democracy. Introduction: The Sceptre and the Spectre. Great-Britain: Palgrave Macmillan
Atwell Seate, A., & Mastro, D. (2015). Media’s influence on immigration attitudes: An ingroup threat theory approach. Communication Manographs, 83(2), 194-213. doi:10.1080/03637751.2015.1068433
Bale, T., Green-Pedersen, C., Krouwel, A., Luther, K. R., & Sitter, N. (2010). If you can’t beat them, join them? Explaining social democratic responses to the challenge from the populist radical right in Western Europe. Political Studies 58(3), 410–426. doi:10.1111/j.1467-9248.2009.00783.x
Blassnig S., & Wirz, D. (2019). Populist and popular: an experiment on the drivers of user reactions to populist posts on Facebook. Social Media + Society, 5(4), 1-12. doi:10.1177/2056305119890062
Bos, L., Lefevere, J. M., Thijssen R, & Sheets, P. (2017). The impact of mediated party issue strategies on electoral support. Party Politics, 23(6), 760-771.
doi:10.1177/1354068815626603
Bos, L., Schemer, C., Corbu, N., Hameleers, M., Andreadis, I., Schulz, A., Schmuck, D., Reinemann, C., & Fazwi, N. (2019). The effects of populism as a social identity frame
on persuasion and mobilisation: Evidence from a 15-country experiment. European Journal of Political Research, 59(1), 3-24. doi:10.1111/1475-6765.12334
Bos, L., van der Brug, W., & de Vreese, C. H. (2011). How the media shape perceptions of right-wing populist leaders. Political Communication,28(2), 182-206.
doi:10.1080/10584609.2011.564605
Bos, L., van Der Brug, W., & De Vreese, C. H. (2013). An experimental test of the impact of style and rhetoric on the perception of right-wing populist and mainstream party leaders. Acta Politica, 48(2), 192-208. doi:10.1057/ap.2012.27
Budge, I. (2015). Issue emphases, saliency theory and issue ownership: a historical and conceptual analysis. Western European Politics, 38(4), 761-777.
doi:10.1080.01402382.2015.1039374
Budge, I., & Farlie, D. (1983). Explaining and predicting elections: Issue effects and party strategies in twenty-three democracies. London: George Allen & Unwin
Canovan, M. (2002). Taking politics to the people: Populism as the ideology of democracy In Mény Y., Surel Y. (eds), Democracies as the populist challenge (25-44). London: Palgrave Macmillan.
de Vreese, C. H., Esser, F., Aalberg, T., Reinemann, C., & Stanyer, J. (2018). Populism as an expression of political communication content and style: A new perspective. The international Journal of Press/Politics 23(4), 423-438.
doi:10.1177/1940161218790035
Dowding, K. (2016). Power and persuasion. Political Studies, 64(1), 4-18. doi:10.1177%2F0032321715614848
Elchardus, M., & Spruyt B. (2016). Populism, persistent republicanism and diclinism: An empirical analysis of populism as a thin ideology. Government and Opposition 51(1), 111–133. doi:10.1017/gov.2014.27.
Engesser, S., Fawzi, N., & Larsson, A. O. (2017). Populist online communication:
introduction to the special issue. Information, Communication & Society, 20(9), 1279-1292. doi:10.1080/1369118X.2017.1328525
Ernst, N., Esser, F., Blassnig, S., & Engesser, S. (2018). Favourable opportunity structures for populist communication: comparing different types of politicians and issues on social media, television and the press. The International Journal of Press/Politics, 00(0), 1-24. doi:10.1177/1940161218819430
FvD, (n.d.). Standpunten. Retrieved at 09-11-2019 via: https://forumvoordemocratie.nl/standpunten
GroenLinks, (n.d.). Goed onderwijs voor iedereen. Retrieved at 27-01-2020 via: https://groenlinks.nl/standpunten/goed-onderwijs-voor-iedereen
Guterres, A. (25 October 2019). Opening remarks at press encounter. United Nations
Secretary-General. Retrieved via: https://www.un.org/sg/en/content/sg/speeches/2019-10-25/remarks-press-encounter
Hameleers, M., Bos, L., & de Vreese, C. H. (2017). “They did it: The effects of emotionalized blame attribution in populist communication.” Communication Research 44(6), 870– 900. doi:10.1177/0093650216644026
Hameleers, M., Bos, L., & de Vreese, C. H. (2018). Start spreading the news: a comparative experiment on the effects of populist communication on political engagement in sixteen European countries. The International Journal of Press/Politics 23(4), 517-538. doi:10.1177/1940161218786786
Hameleers, M., & Schmuck D. (2017). It’s us against them: a comparative experiment on the effects of populist messages communicated via social media. Information,
Hogg, M., & Reid, S. (2006). Social identity, self-categorization, and the communication of group norms. Communication Theory, 16(1), 7-30.
doi:10.1111/j.1468-2885.2006.00003.x
Holian, D. (2004). He’s stealing my issues! Clinton’s crime rhetoric and the dynamics of issue ownership. Political Behaviour, 26(2), 95-124.
doi:10.1023/B:POBE.0000035959.3556
Hong, S., & Nadler, D. (2011). Does the early bird move the polls?: the sue of the social media tool ‘Twitter’ by U.S. politicians and its impact on public opinion. Proceedings of the 12th Annual International Digital Government Research Conference: Digital Government Innovation in Challenging Times, 182-186.
doi:10.1145/2037556.2037583
Householder, E. E., & LaMarre, H. L. (2014). Facebook politics: Towards a process for achieving political source credibility through social media. Journal of Information Technology & Politics, 11(4), 368-382. doi:10.1080/19331681.2014.951753
Jagers, J., & Walgrave, S. (2007). Populism as political communication style: An empirical study of political parties’ discourse in Belgium. European Journal of Political Research, 46, 319–345. doi:10.1111/j.1475-6765.2006.00690.x
Kessel, S. (2011). Explaining the electoral performance of populist parties: the Netherlands as a case study. Perspectives on European Politics and Society, 12(1), 68-88.
doi:10.1080/15705854.2011.546148
Kramer, B. (2017). Populist online practices: the function of the Internet in right-wing populism. Information, Communication & Journalism, 20(9), 1293-1309. doi:10.1080/1369118X.2017.1328520
Kwak, N., Lane, D., Lee, S., Liang, F., Weeks, B. (2018). From persuasion to deliberation: do experiences of online political persuasion facilitate dialogic openness? Communication Research, 1-23, doi:10.1177/0093650218793050
Lecheler, S., Schuck, A. R., & de Vreese, C. H. (2013). Dealing with feelings: Positive and negative discrete emotions as mediators of news framing effects. Communications 38(2), 189-209. doi:10.1515/commun-2013-0011
Matland, R. E., & Murray, G. R. (2013). An experimental test for “backlash” against social pressure techniques used to mobilize voters. American Politics Research, 41(3), 359-386. doi:10.1177/1532673X12463423
Mols, F. (2012). “What makes a arame persuasive? Lessons from social identity theory.” A Journal of Research, Debate and Practice 8(3), 29–45.
doi:10.1332/174426412X654059.
Mudde, C. (2000). The Ideology of the Extreme Right. Manchester: Manchester University Press.
Mudde, C. (2007). Populist radical right parties in Europe. Cambridge University Press. doi:10.1017/CBO9780511492037
Mudde, C. (2013). “Three decades of populist radical right parties in Western Europe: So What?” European Journal of Political Research 52(1), 1-9.
doi:10.1111/j.14756765.2012.02065.x
Mudde, C. & Taggart, P. (2019). The Populist: An Overview of Populist, Far Right, Far Left and Eurosceptic Parties in Europe. Retrieved via: http://www.popu-list.org.
PVV, (n.d.). Concept verkiezingsprogramma 2017-2021. Retrieved at 09-11-2019 via: https://www.pvv.nl/visie.html
Reinemann, C., Aalberg, T., Esser, F., Strömbäck, J., & de Vreese, C. H. (2017). Populist political communication: Toward a model of its causes, forms, and effects. In Aalberg,
T., Esser, F., Reinemann, C., Strömbäck, J., & de Vreese, C. H. (Eds.), Populist Political Communication in Europe (pp. 12-25). (Routledge Research in
Communication Studies; Vol. 1). New York: Routledge.
Rocklage, M., Derek, D., & Nordgren, L. (2018). Persuasion, emotion and language: the intent to persuade transforms langue via emotionality. Psychological Science, 29(5), 749-760. doi:10.1177/0956797617744797
Rocklage, M., & Fazio, R. (2016). On the dominance of attitude emotionality. Personality and Social Psychology Bulleting, 42(2), 259-270. doi:10.1177/0146167215623273 Rocklage, M, Rucker, D., & Nordgren, L. (2017). The evaluative lexion 2.0: The
measurement of emotionality, extremity, and valence in langue. Behavior Research Methods, 50(4), 1327-1344. doi:10.3758/s13428-017-0975-6
Rooduijn, M., de Lange, S., & van der Brug, W. (2014). A populist Zeitgeist? Programmatic contagion by populist parties in Western Europe. Party Politics, 20(4), 563-575. doi:10.1177/1354068811436065
Rooduijn, M., van Kessel, S., Froio, C., Pirro, A., de Lange, S., Halikiopoulou, D., Lewis, P., Mudde, C. & Taggart, P. (2019). The Populist: An Overview of Populist, Far Right, Far Left and Eurosceptic Parties in Europe. http://www.popu-list.org.
Schedler, A. (1996) Anti-political establishment parties. Party Politics 2(3): 291–312. doi:10.1177/1354068896002003001
Scheufele, D., & Tewksbury, D. (2007). Framing, agenda setting, and priming: The evolution of the three media effect models. Journal of Communication, 57(1), 9-20.
doi:10.1111/j.0021-9916.2007.00326.x
Schmitt-Beck. D., & Farrell, D. (2002). Do political campaigns matter? Campaign effects in elections and referendums. London: Routledge
Schmuck, D., & Matthes, J. (2017). Effects of economic and symbolic threat appeals of right-wing populist advertising on anti-immigration attitudes. The impact of textual and visual appeals. Political Communication 34(4), 604-626.
doi:10.1080/10584609.2017.1316807
Sheets, P., Bos, L., & Boomgaarden, H. G. (2015). Media cues and citizens support for right-wing populist parties. International Journal of Public Opinion Research, 28(3), 307-330. doi:10.1093/ijpor/edv014
SP (n.d.-a). Heel de mens – kernvisie. Retrieved at 09-11-2019 via: https://www.sp.nl/publicatie/heel-de-mens/kernvisie
SP (n.d.-b). Onderwijs en cultuur. Retrieved at 27-01-2020 via: https://www.sp.nl/thema/onderwijs-en-cultuur
Stier, S., Posch, L., Bleier, A., & Strohmaier, M. (2017). When populists become popular: comparing Facebook use by the right-wing movement Pegida and German political parties. Information, Communication & Society, 20(9), 1365-1388.
doi:10.1080/1369118X.2017.1328519
Taggart, P. (2000). Populism. Buckingham: Open University Press
Taggart, P. (2006). Populism and representative politics contemporary Europe. Journal of Political Ideologies, 9(3), 269-288. doi:10.1080/1356931042000263528
van Spanje, J., & de Vreese, C. H. (2015). ‘The good, the bad and the voter: The impact of hate speech prosecution of a politician on electoral support for his party’. Party Politics 21(1), 115-130. doi:10.1177/1354068812472553
Walgrave, S., Lefevere, J., & Tresch, A. (2012). The associative dimension of issue ownership. Public Opinion Quarterly, 76(4), 771-782. doi:10.1093/poq/nfs023
Weyland, K. (2001). Clarifying a contested concept: populism in the study of Latin American politics. Comparative Politics, 34(1), 1-22. doi:10.2307/422412
Wirz, D., Wettstein, M., Schulz, A., Müller, P., Schemer, C., Ernst, N., Esser, F., & Wirth, W. (2018). The effects of right-wing populist communication on emotions and cognitions towards immigrants. The International Journal of Press/Politics, 23(4), 496-516. doi:10.1177/1940161218788956
Appendix Appendix I
Pre-test
In order to make sure that the stimuli were reliable a pre-test was conducted for each of the four conditions. First of all, a factor analysis was conducted to measure if the messages were perceived as populist or not. This was assessed by five items on a 7-point scale. The items tapped how much participants (dis)agreed with the populist statements. The items were (1) ‘The message negatively presents the political elite’, (2) ‘The message is populist’, (3) ‘The government favours the capitalists’, (4) ‘The Dutch children are victims of the teachers’ shortage’ and (5) ‘The message blames the government for the problems in the education sector’.
For condition 1 (populist message from the SP), the an Exploratory factor analysis (principal axis with Varix) showed that there are two factors (eigenvalue factor 1: 1.91; eigenvalue factor 2: 1.73). Together they explained 72.6% of the variance. The factor analysis showed a good reliability of the scale with a Cronbach’s alpha of 0.69.
For condition 2 (neutral message from the SP), an Exploratory factor analysis (principal axis with Varimax) showed two factors (eigenvalue factor 1: 2.96; eigenvalue factor 2: 1.41). Together they explained 87.4% of the variance. The factor analysis showed a good reliability of the scale with a Cronbach’s alpha of 0.70.
For condition 3 (populist message from GroenLinks), an Exploratory factor analysis (principal axis with Varimax) showed two factors (eigenvalue factor 1: 3.74; eigenvalue factor 2: 1.12). Together they explained 74.8% of the variance. The factor analysis showed a good reliability of the scale with a Cronbach’s alpha of 0.75.
For condition 4 (neutral message from GroenLinks), an Exploratory factor analysis (principal axis with Varimax) showed two factors (eigenvalue factor 1: 2.95; eigenvalue factor 2: 1.49).
Together they explained 58.9% of the variance. The factor analysis showed a good reliability of the scale with a Cronbach’s alpha of 0.83. A scale was set up for each of the four
conditions. Table 1. shows that the messages where perceived as how they were intended: the neutral message as neutral and the populist messages as populist.
Table 1. How the messages were perceived in the pre-test
Conditions M SD
(1) SP populist 5.71 0.82
(2) SP neutral 3.29 1.10
(3) GroenLinks populist 5.43 1.35 (4) GroenLinks neutral 3.48 1.16
Furthermore, the credibility of the message was assessed via the question ‘How credible did you find the message?’ on a 7-point scale. Table 2. shows that the credibility of the messages was good.
Table 2. Credibility of the message
Conditions M SD
(1) SP populist 4.57 0.98
(2) SP neutral 5.43 0.98
(3) GroenLinks populist 4.00 2.01 (4) GroenLinks neutral 5.00 1.00
Appendix II Tables and Figures
Table 3. Descriptives of the sample
Descriptives M SD Minimum Maximum
Age 41.91 18.60 19 74
Gender (%):
Male 37.1%
Female 62.9%
Political identification 5.08 1.78 1 ultimate left-wing 10 ultimate right-wing General interest in
politics
4.98 1.28 1 not interested at all 7 very interested
Identification of
importance of the topics:
1 not interested at all 7 very interested
Defence 4.41 1.47
Education 6.23 0.81
Reside 5.75 0.69
Level of persuasion 5.00 1.00 1 totally disagree 7 totally agree Probability to vote 3.31 1.75 1 not likely at all 7 very likely Feeling positive/negative 3.77 1.37 1 negative 7 positive
Evoked emotions: 1 describes my
feeling totally not
7 describes my feeling completely Anger 4.27 1.47 Frustration 4.47 1.39 Hope 3.50 1.39 Sadness 3.70 1.44 No emotions 3.60 1.62
Figure 2 . Education level of the sample 24,1 18,8 17,1 10 5,9 2,9 2,4 1,8 1,2 9,4 1,2
Voting Behaviour
D66 24.1% GroenLinks 18.8% VVD 17.1%PvdA 10.0% CDA 5.9% ChristenUnie 2.9%
FvD 2.4% PVV 1.8% 50Plus 1.2%
No idea 9.4% Would not go voting 1.2% 35,9 19,4 15,9 9,4 8,8 5,3 1,8 0,62,9
Education
HBO 35.9% MBO 19.4% Master Degree 15.9%
Bachelor's Degree 9.4% HAVO 8.8% VWO 5.3%
VMBO 1.8% Primary School 0.6% Differently 2.9%
Table 4. Manipulation check
Table
5. Manipulation check type of partyStatements Populist condition Neutral condition T test M SD M SD Message blames the
government for the problems in the education system
t = 2.33, p = .021 5.71 1.31 5.21 1.41
The Dutch children are portrayed as a victim of the problems in the education system
t = 1.70, p = .091 5.48 1.46 5.06 1.62
Capitalists are the governments favour
t = 6.35, p = .000 4.97 1.53 3.26 1.84
The message is negative about the political elite.
t = 6.90, p = .000 4.84 1.53 3.12 1.60
The message is populist.
t = 5.01, p = .000 4.64 1.50 3.32 1.75
Party t Group SP Group GroenLinks
PvdA t = -0.58 p = .565 M = 2.89, SD = 2.09 M = 2.70 , SD = 2.07 SP t = 8.89, p = .000 M = 5.44, SD = 2.04 M = 2.59, SD = 2.04 GroenLinks t = -9.93, p = .000 M = 2.78 SD = 2.04 M = 5.81 SD = 1.80
Table 6. Manipulation check conditions
Table 7. Dependent variables evoked emotions H2 Credibility of stimulus material M SD SP populist 4.80 1.49 SP neutral 4.90 1.46 GL populist 4.97 1.70 GL neutral 5.07 1.40 F(3,165) = 0.23 p = .873 Type of emotions
Levene’s test T test Populist message Neutral message M SD M SD Anger 1.93 p = .167 -1.09, p = .278 4.15 1.56 4.40 1.37 Frustration 1.96 p = .163 -1.13 p = .261 4.32 1.50 4.57 1.28 Hope 0.21 p = .646 -1.12 p = .266 3.40 1.30 3.56 1.44 Sadness 0.01 p = .934 -0.07 p = .948 3.69 1.44 3.71 1.46
Table 8. Dependent variable evoked emotion RQ2a
Table 9. Regression model dependent variable evoked emotion RQ2b Type of
emotions
Levene’s test T test Populist SP Populist GL
M SD M SD Anger 0.13 p = .722 -0.43 p = .667 4.08 1.57 4.23 1.56 Frustration 1.02 p = .316 -0.55 p = .581 4.23 1.63 4.42 1.37 Hope 3.52 p = .064 -1.36 p = .179 3.21 1.42 3.61 1.15 Sadness 0.15 p = .699 -0.11 p = .914 3.58 1.48 3.56 1.43 Type of emotions b t p Anger Populist message Populist party Message*Party -0.41 -0.48 0.32 -1.27 -0.15 0.70 .208 .138 .485 F test p 1.21 .309 R2 2.2% Frustration Populist message Populist party Message*Party F test p R2 -0.27 -0.24 0.05 0.74 .528 1.4% -0.86 -0.78 0.11 .389 .436 .910
Hope Populist message Populist party Message*Party F test p R2 -0.47 -0.87 0.47 3.89 .010 0.7% -1.55 -2.94 1.11 .122 .004 .270 Sadness Populist message Populist party Message*Party F test p R2 -0.17 -0.36 -0.32 0.42 .742 0.8%. -0.06 -0.12 0.10 .601 .269 .487
Appendix III Stimulus material
Figure 4. Populist text from populist party SP
Appendix IV
Figure .. non-populist text from non-populist party GroenLinks
Figure 6. Populist text from non-populist party GroenLinks
