Projects

Project description:

This project examines how the emotional intensity of public opinion signals affects the public opinion estimations of politicians. Numerous works, including previous studies that were conducted as part of this research centre, show that politicians tend to follow their public opinion estimation in their voting or communication. This makes the accuracy of politicians’ public opinion estimates highly relevant. The core finding, though, is that politicians do not hold accurate perceptions. Their perception is not only inaccurate but also biased. There is a systematic right-wing bias in these perceptions; politicians think the public holds more right-wing policy preferences than it does. This bias in perceptions can lead to policies that are more right-wing than the public expects them to be and this may fuel, in the next step, discontent of citizens with their political leaders.

Where does the right-wing bias in perception come from? The innovative hypothesis underlying this project is that the right-wing bias is due to how right-wing citizens express their political preferences compared to left-wing citizens. Our key expectation is that right-wing citizens signal their preferences to politicians in a more emotionally intense way, also exhibiting different types of emotions, and that this leads to the right-wing bias in politicians’ public opinion perceptions.

To examine how the emotional intensity of public opinion signals affects the public opinion estimations of politicians, the project draws on interviews, surveys, and survey-experiments applied comparatively in eight countries (Australia, Belgium, Canada, Czech Republic, Israel, Portugal, Sweden, and Switzerland, see also POLPOP project), as well as lab experiments grasping politicians’ physiological reactions to public opinion messages.

Researchers:

• Stefaan Walgrave (Leader)
• ​Karolien Poels (Co-leader)​
  
• Julie Sevenans (Post-doctoral researcher)
• Karolin Soontjens (Post-doctoral researcher)
• ​Konrad Rudnicki (Post-doctoral researcher)
  

Project description:

This project analyses the interactions between politicians and citizens on social media, thereby focusing on real-life interactions between the two main democratic actors. The plan is to examine the communication of politicians on social media, and the interactions with the audience, longitudinally over time (2010-2025) in Belgium through an ambitious, longitudinal content analysis. The advantage of such an approach is that we study real-world behaviour and not just self-reported beliefs, attitudes, or (intended) behaviour. To put the Belgian case into perspective, we add a comparative component by analysing social media content of politicians in eight countries (Australia, Belgium, Canada, Czech Republic, Israel, Portugal, Sweden, and Switzerland) within a shorter period (2023-2024)―advances in automatic text classification make this feasible.

Regarding politicians’ communication, we focus on two aspects: content and style. First, social media allow politicians to promote the topics they work on, and potentially influence the public and/or media agenda. For instance, we expect that the issue agenda of politicians on social media follows a combination of party (issue ownership) and media logic (salient issues in the news). This project aims to give insight into which logic is dominant for which types of politicians, and how this changes over time. Second, we study the communication style of politicians. To what extent do politicians use negative, populist, and uncivil language when communicating with the public? In line with one of the major innovations of INCIPOL, we devote special attention to the emotions conveyed in political text.

From the side of the public, this project studies which messages in terms of both issue and style, lead to the most engagement (likes, comments, shares). Does emotional politician communication lead to emotional citizen reactions? However, more innovatively, we also examine to what extent these reactions of the public, in turn, influence what and how politicians post on social media. Do, for instance, more reactions of followers on messages about a contested issue such as migration spur politicians to focus even more on this topic? Or does more incivility and emotion in the comments of the audience inspire politicians to adopt a more uncivil and emotional style? In other words, does the communication of citizens influence the communicative agenda and style of politicians? The longitudinal design of this study allows us to examine when and how this potential feedback loop from the (online) public to politicians works.

Researchers:

• Peter Van Aelst (Leader)
• Stefaan Walgrave (Co-leader)
• ​Jonas Lefevere (Post-doctoral researcher)
• ​Željko Poljak (Post-doctoral researcher)
  
• ​August De Mulder (Post-doctoral researcher)
  
• ​Evelien Willems (Post-doctoral researcher)​

Project description:

Hate messages against politicians on social media are thriving. In our own interviews with politicians, many say that the toxic nature of social media is a constant burden. Other work looking at the comments to tweets by elected representatives in the US and UK reveals that hate messages are frequent and that some politicians are targeted more than others: especially female politicians and politicians of colour are under fire. Although it is never acceptable to target individual politicians with real hate such as racism, sexism, or encouraging violence towards a person, hateful messages can reflect citizens’ real worries. Hate is typically evoked when people feel threatened and want to protect themselves and their nears. Although it is only human for politicians to be affected by hate, one could argue that, as representatives, politicians must take the sentiments of citizens seriously, even if those sentiments are expressed in an uncivil or even downright unacceptable way. Simply ignoring these messages is, from a democratic perspective, not the best coping strategy.

This underscores the relevance of studying how politicians can deal with (different types of) hate messages and filter out valuable policy critique. Hence, the double research question of this project: (1) How do politicians react to and interact with online hate messages, and how do they deal with or cope with online hate messages in relation to their own well-being; (2) Can we develop strategies to help politicians deal with hate messages and to turn hate into a valuable public opinion signal?

To answer these questions, we draw on content analysis to examine the incoming hate on Belgian politicians’ social media and to compare across individuals and parties. This is complemented with interviews with politicians to tap into how much and what kind of online hate politicians receive, as well as how they cope with hate and how it affects them. Furthermore, we will do survey-experiments to test the causal effect of hate speech, but also to explore whether in some circumstances, cognitive reappraisal of online hate messages as reflecting citizens’ emotions with sincere intentions expressed in an extreme, helpless, frustrated way, can make politicians feel less personally affected and lead to a higher self-efficacy to cope and respond.

Researchers:

• ​Karolien Poels (Leader)​
  
• ​Stefaan Walgrave (Co-leader)
• ​Konrad Rudnicki (Post-doctoral researchers)
• Julie Sevenans (Post-doctoral researchers)
• Ine Goovaerts (Post-doctoral researchers)

Project description:

Recent research focused on how politicians think about public opinion, and how they convey messages to the wider public or a specific constituency. However, few studies looked into the connection of both. This is the research question this project aims to tackle, innovatively connecting how politicians read public opinion signals and how it shapes their subsequent communication.

The project draws on several methods. Content analysis of social media will serve to examine actual politicians’ communication in Belgium. Interviews with politicians in eight countries (Australia, Belgium, Canada, Czech Republic, Israel, Portugal, Sweden, and Switzerland) will serve to qualitatively dig up the dimensions of communication that may be affected and disentangle the features of more or less ‘assertive’ speech. Moreover, relying on surveys with politicians, closed questions will be asked about how public opinion perceptions matter for whether and how politicians communicate. These surveys will also include survey-experiments wherein different variants of public opinion are encapsulated in a series of vignettes and politicians are queried about their likely communicative follow-up.

Researchers:

• Stefaan Walgrave (Leader)
• ​Peter Van Aelst (Co-leader)
• Jonas Lefevere (Post-doctoral researcher)

Project description:

This project addresses the emotional responses of citizens to political rhetoric based on physiological experiments. Recent scholarship increasingly acknowledges the role of emotions in shaping the effects of political communication and the limitations of self-reporting in capturing emotional responses. Physiological responses are proposed as a more reliable, valid, and direct metric of emotional responses. Previous work shows the absence of a strong correlation between physiological responses and self-reported data in the context of politics and political communication. The limited work that exists suggests that emotions measured physiologically predict outcomes such as evaluations of policies or issue stances. This project pursues that track further and examines unexplored avenues about the intermediary role of emotional reactions of citizens to politicians’ communication leading to cognitive, attitudinal, and behavioural effects.

From 2026 onwards, we plan a series of physiological experiments with Belgian citizens in the Antwerp Social Lab―a state-of-the-art infrastructure possessing all the necessary tools. We dive deep into emotional responses through direct physiological recordings while citizens are exposed to political communication, and we will compare and complement these responses with self-reported, later-recorded emotions. In addition, we will also look into self-reported outcome variables such as beliefs about reality, attitudes about parties, policies, and issues, and (intended) behaviour.

Researchers:

• ​Karolien Poels (Leader)​
  
• Stefaan Walgrave (Co-leader)
• ​Konrad Rudnicki (Post-doctoral researcher)​