Periodic Reporting for period 4 - POLEMIC (Politics and Emotions Investigated Comparatively)
Reporting period: 2022-10-01 to 2024-03-31
In 2016 with the surprise election of Trump and the Brexit referendum, there was widespread concern about how emotional and negative the rhetoric of politicians had become. Also, increasingly, media pointed to emotions such as anger and anxiety to explain what looked like a general, global democratic revolt against the establishment. The POLEMIC project was borne out of these concerns.
The key results are:
Politics is neither more emotional nor more negative
Citizens want more positive emotions
Emotion is deeply involved in political decision-making
Emotions about politics have a complex structure
Some people are more emotional about politics than others
The next section explains these conclusions in more detail.
The key results are:
Politics is neither more emotional nor more negative
Citizens want more positive emotions
Emotion is deeply involved in political decision-making
Emotions about politics have a complex structure
Some people are more emotional about politics than others
The next section explains these conclusions in more detail.
Politics is neither more emotional nor more negative
We systematically tracked the tone in speeches in 7 parliaments (UK, Netherlands, Austria, Sweden, Denmark, Germany and Spain) over 30 years (Pipal, Bakker, et al., 2024). We find no systematic trend towards more positive or more negative tone in these speeches (Figure 1a). In some countries speeches became more positive, in others more negative. Also, factors such as ideology, gender, economic and cultural trends do not explain the variation in tone (Pipal, Bakker, et al., 2024). We do find that government politicians strike a more positive tone than opposition politicians. Also, we have some tentative evidence that the personality of politicians shapes the tone of their speeches (Pipal, 2024a).
The method we used is explained in Pipal, Schoonvelde, and Schumacher (2024). We also created an R-package based on this method for the research community to use (Pipal, 2024b).
Citizens want more positive emotions
In two representative surveys we asked participants to indicate whether they want their party leader to use more or less of a series of positive and negative emotions. As Figure 1B shows participants favor more positive emotions and fewer negative emotions. In two different surveys we find that approximately 7 out of 10 participants agree that positive emotions should have a role in politics, whereas negative emotions should not.
Also using an experimental task shows that politicians showing positive facial expressions or using positive tone receive better evaluations than politicians with negative expressions or using negative tone (Homan & Schumacher, 2023), see Figure 1D. We ran a total of 10 experiments in the US, Netherlands, Greece and Poland using the same task and finding the same result. Figure 1D summarizes the results of the first 4 studies.
Emotion is deeply involved in political decision-making
The experiment discussed in the previous paragraph demonstrates that positive emotions make political messages more persuasive. More so than negative or neutral tone and expressions do.
People process arguments they agree or disagree with differently. Using a lab-in-the-field design we found that the corrugator muscle - associated with negative affect - is activated when people process arguments they disagree with (Bakker et al., 2021), see Figure 1F. This physiological activity was also associated with post-treatment attitudinal change.
Second, people process faces of inparty and outparty politicians differently. For example, outparty politicians faces evoke more activity of facial muscles, in particular the zygomaticus muscle (Homan et al., 2023). This muscle is associated with positive affect. It seems outparty politicians amuse us or that we feel contempt for them, see Figure 1C. Also there is a difference in neural activity: processes associated with attention (measured by so-called alpha rhythm and perspective-taking (mu rhythm) (Homan et al., 2024) are more strongly activated in the outparty condition compared to the inparty condition. Finally, a lab experiment with a slightly different design we show that people have instantaneous disgust responses (levator labii muscle) to outparty leaders (Bakker, Schumacher, & Homan, 2020).
A final piece of evidence here is that packaging emotions into political messages does not always work, even when there is no politician involved and people are neutral towards the message. In a series of survey experiments we designed messages that should elicit anger or anxiety. We did this by manipulating descriptions of situations following the appraisal model. Our participants did report more anger in the anger condition, but they also report fear, both in the fear condition and in the other conditions. We were able to manipulate anger but not fear (Rebasso, 2023). This led to the broader question how distinctive emotions are in the context of politics.
Emotions about politics have a complex structure
Emotions in the political context are much harder to identify than emotions in general. In two surveys we asked survey participants to describe an emotional event and to label their emotion. Then ask them questions how they appraised that situation. This we used to predict the emotion people had indicated with relatively high accuracy. Yet, we asked half of our participants to think about an emotional event related to politics. In this condition the accuracy of our predictions fell dramatically. In fact, we could no longer reliably distinguish between negative emotions such as anger, fear and sadness (Rebasso, 2023). Also when we compare physiological activity related to emotions with self-reports of emotions we fail to find a connection. This is both the case for neutral (Bakker, Schumacher, Gothreau, & Arceneaux, 2020) and political stimuli (Bakker, Schumacher, & Homan, 2020; Bakker et al., 2021; Homan, 2024; Homan et al., 2023). This underlines the disconnection between unconscious and conscious processes (Arceneaux et al., in-press; Bakker et al., 2021; Schumacher et al., 2024). It raises the question whether we can be better in touch with the more unconscious, emotional processes.
Some people are more emotional about politics than others
Figure 1E shows characteristics of people who feel more and less emotions about politics. The more people are interested and feel they have political efficacy, the stronger the emotions they experience when they think about politics (Rebasso, 2023). These results are consistent across Dutch and American data. Knowledge, ideology and confidence are less consistently related to self-reported emotions. But there is also a big gender gap with women reporting more negative emotions and men more positive emotions.
Conclusion
Our results suggest no need for alarmism about emotions. There is not a singular trend towards increasing negativity in politics. Tone and facial expressions affect people but the effect sizes we find are modest and also much lower than the effects of agreement with a politician on substantive terms. More worrying is perhaps the lack of connection between people’s conscious experience of their emotions, the emotion labels they use, and the unconscious processes that drive emotions, information processing, and political behaviours (Homan, 2024; Rebasso, 2023).
Should politics be strictly rational, and emotions be banned from the domain? Although some have suggested this, turning off emotions means turning off the brain. That is, affective and cognitive processes are deeply interconnected. As such, the question ought to be: how can we use emotions to be a positive force for democracy?
We systematically tracked the tone in speeches in 7 parliaments (UK, Netherlands, Austria, Sweden, Denmark, Germany and Spain) over 30 years (Pipal, Bakker, et al., 2024). We find no systematic trend towards more positive or more negative tone in these speeches (Figure 1a). In some countries speeches became more positive, in others more negative. Also, factors such as ideology, gender, economic and cultural trends do not explain the variation in tone (Pipal, Bakker, et al., 2024). We do find that government politicians strike a more positive tone than opposition politicians. Also, we have some tentative evidence that the personality of politicians shapes the tone of their speeches (Pipal, 2024a).
The method we used is explained in Pipal, Schoonvelde, and Schumacher (2024). We also created an R-package based on this method for the research community to use (Pipal, 2024b).
Citizens want more positive emotions
In two representative surveys we asked participants to indicate whether they want their party leader to use more or less of a series of positive and negative emotions. As Figure 1B shows participants favor more positive emotions and fewer negative emotions. In two different surveys we find that approximately 7 out of 10 participants agree that positive emotions should have a role in politics, whereas negative emotions should not.
Also using an experimental task shows that politicians showing positive facial expressions or using positive tone receive better evaluations than politicians with negative expressions or using negative tone (Homan & Schumacher, 2023), see Figure 1D. We ran a total of 10 experiments in the US, Netherlands, Greece and Poland using the same task and finding the same result. Figure 1D summarizes the results of the first 4 studies.
Emotion is deeply involved in political decision-making
The experiment discussed in the previous paragraph demonstrates that positive emotions make political messages more persuasive. More so than negative or neutral tone and expressions do.
People process arguments they agree or disagree with differently. Using a lab-in-the-field design we found that the corrugator muscle - associated with negative affect - is activated when people process arguments they disagree with (Bakker et al., 2021), see Figure 1F. This physiological activity was also associated with post-treatment attitudinal change.
Second, people process faces of inparty and outparty politicians differently. For example, outparty politicians faces evoke more activity of facial muscles, in particular the zygomaticus muscle (Homan et al., 2023). This muscle is associated with positive affect. It seems outparty politicians amuse us or that we feel contempt for them, see Figure 1C. Also there is a difference in neural activity: processes associated with attention (measured by so-called alpha rhythm and perspective-taking (mu rhythm) (Homan et al., 2024) are more strongly activated in the outparty condition compared to the inparty condition. Finally, a lab experiment with a slightly different design we show that people have instantaneous disgust responses (levator labii muscle) to outparty leaders (Bakker, Schumacher, & Homan, 2020).
A final piece of evidence here is that packaging emotions into political messages does not always work, even when there is no politician involved and people are neutral towards the message. In a series of survey experiments we designed messages that should elicit anger or anxiety. We did this by manipulating descriptions of situations following the appraisal model. Our participants did report more anger in the anger condition, but they also report fear, both in the fear condition and in the other conditions. We were able to manipulate anger but not fear (Rebasso, 2023). This led to the broader question how distinctive emotions are in the context of politics.
Emotions about politics have a complex structure
Emotions in the political context are much harder to identify than emotions in general. In two surveys we asked survey participants to describe an emotional event and to label their emotion. Then ask them questions how they appraised that situation. This we used to predict the emotion people had indicated with relatively high accuracy. Yet, we asked half of our participants to think about an emotional event related to politics. In this condition the accuracy of our predictions fell dramatically. In fact, we could no longer reliably distinguish between negative emotions such as anger, fear and sadness (Rebasso, 2023). Also when we compare physiological activity related to emotions with self-reports of emotions we fail to find a connection. This is both the case for neutral (Bakker, Schumacher, Gothreau, & Arceneaux, 2020) and political stimuli (Bakker, Schumacher, & Homan, 2020; Bakker et al., 2021; Homan, 2024; Homan et al., 2023). This underlines the disconnection between unconscious and conscious processes (Arceneaux et al., in-press; Bakker et al., 2021; Schumacher et al., 2024). It raises the question whether we can be better in touch with the more unconscious, emotional processes.
Some people are more emotional about politics than others
Figure 1E shows characteristics of people who feel more and less emotions about politics. The more people are interested and feel they have political efficacy, the stronger the emotions they experience when they think about politics (Rebasso, 2023). These results are consistent across Dutch and American data. Knowledge, ideology and confidence are less consistently related to self-reported emotions. But there is also a big gender gap with women reporting more negative emotions and men more positive emotions.
Conclusion
Our results suggest no need for alarmism about emotions. There is not a singular trend towards increasing negativity in politics. Tone and facial expressions affect people but the effect sizes we find are modest and also much lower than the effects of agreement with a politician on substantive terms. More worrying is perhaps the lack of connection between people’s conscious experience of their emotions, the emotion labels they use, and the unconscious processes that drive emotions, information processing, and political behaviours (Homan, 2024; Rebasso, 2023).
Should politics be strictly rational, and emotions be banned from the domain? Although some have suggested this, turning off emotions means turning off the brain. That is, affective and cognitive processes are deeply interconnected. As such, the question ought to be: how can we use emotions to be a positive force for democracy?