Master’s Thesis
From Heads to Hearts – Emotions in Political Communication:
How do appraisals elicit emotions and subsequently affect attitudes and behavioral intentions?
An experimental study in the context of the new German immigration law
Patricia Weiß
Student ID: 11832827
Master’s programme Communication Science
Supervisor: Lukas Otto
Word Count: 89741
Date of completion: 01. February 2019
1 Abstract
Given the political climate, immigration is a contemporary political debate in Germany, Europe, and the United States infused with emotions. The experience of emotion is closely associated with the individual’s appraisal of its environment along with several cognitive appraisals (pleasantness, certainty, control). Although extensive research in psychology has proven that emotions are extracted from appraisals and subsequently affect attitudes and behavioral intentions, it has surprisingly few empirical foundations in political communication, to date. Studying appraisals is attractive to communication scholars because it allows them to understand what the media consists of, how citizens evaluate messages and acquire information to make informed political decisions. Drawing on appraisal theories, this experimental study is among the first who manipulates the aforementioned appraisal patterns in newspaper articles according to the emotional states of empathy, hope, anger, and fear in the context of the new German immigration law. The results indicate that appraisals are important predictors for emotional arousal; especially the empathy and fear appraisals could elicit the corresponding emotion. Based on these findings anger and fear mediated the effects of appraisals on policy rejection and negative attitudes towards immigration on the German labor market, while hope correlated with policy support, positive attitudes, and approach behavior. Surprisingly, empathy mediated the effect of appraisals on behavioral intentions indicating that a positive (and
vicarious) emotion motivates citizens to engage in politics. With this study, I provide further building blocks for the integration of appraisal dimensions as an explanation for the elicitation of emotions in political communication and the subsequent attitudinal and behavioral outcomes. Also, I extend the evidence of positive emotions and empathy highlighting the importance of it in political communication.
2 Introduction
Emotion matters. Emotions are integral to understanding democratic citizenship because they shape how people perceive their political world (Marcus, 2002). We get involved in politics because we feel compelled by our emotions (Crigler & Just, 2012). Contemporary political events such as the election of Donald Trump, Brexit or the Syrian refugee crisis have prompted widespread emotional response (Romano, 2018). This is also the case for the new German immigration law passed in December 20182. The press has often emotionally portrayed the
regulation, sparking a controversial debate in public (May, 2018). It seems evident that political communication is capable of arousing an emotional response.
Previous studies showed how immigration is presented in the news influences the public’s perceptions, opinions, and attitudes towards immigration mediated by primarily negative emotions (Gadarian & Albertson, 2014; Igartua, Moral-Toranzo, & Fernández, 2011). Public discourse in general and the mentioned studies, in particular, tend to emphasize the costs rather than the benefits of immigration (Brader, Valentino, & Suhay, 2008). Although this reflects the general bias of negativity in the news (Hibbing & Theiss-Morse, 1998), scholars emphasized to study the consequences of positive emotions since they have been neglected in political
communication research (Brader et al., 2011). Positive emotions have more longlasting effects and broaden the individual’s thought and action patterns (Fredrickson, 2001) as well as having a motivating impact (Holm, 2012). Also, they are also likely to shape opinions about immigration (Griskevicius, Shiota, & Neufeld, 2010; Verkuyten, 2004). Further, the rarely examined emotion of empathy could be of interest when studying media effects in an immigration context, as suggested but not tested by Lecheler, Bos, and Vliegenthart (2015).
2 Keeping the aging population and the shrinking workforce in mind, the German economy needs more flexibility
to meet the demand for high- and low-skilled workers. Hence, the purpose of this law is to facilitate the migration of foreign workers. Not only in Germany but also both the United States of America and Great Britain, migration of high and low skilled workers has become an increasingly contentious political issue (Nienaber, 2018).
3 Until now, prior research mainly provides insight on media frames as a rhetorical device of news reporting which can elicit strong emotional responses (Aarøe, 2011; Druckman &
McDermott, 2008; Gross & D’Ambrosio, 2004). For many years, political communication researchers have overlooked the importance of frames on affective experience; however, there is little to no literature available experiencing emotions derived from individuals’ cognitive
appraisal of events. This is surprising since appraisals are one of the dominant explanations in psychology for comprehending distinct patterns of emotions (Lazarus, 1991; Roseman, 1991; Scherer, 1999). To address this gap, this experimental study breaks new ground in political communication. It will transfer the theoretical considerations, expectations, and findings from emotions research in psychology to the arena of political communication. In a first step, the appraisal patterns pleasantness, certainty, and control (Smith & Ellsworth, 1985) will be manipulated according to the discrete emotions3 of hope, anger, fear, and empathy by using
particular news content about the new German immigration law. Examining readers’ appraisals and emotions in response to news story content may have important implications for media communication because “(often small) changes in the presentation of an issue or an event produce (sometimes large) changes” in attitudes and behavior (Chong & Druckman, 2007, p. 104). Consequently, the second part of this study examines the spillover effects of positive and negative emotions on policy support or rejection as well as attitudes and behavioral intentions towards immigration on the labor market in Germany. Thereby, I expect hope, empathy, fear, and anger to have distinct effects due to underlying disparate appraisal patterns.
Theoretical Framework
The theoretical framework of this paper will be based on appraisal theories (Lazarus, 1991), especially appraisal dimensions (Smith & Ellsworth, 1985), the cognitive functional model of
3Those with subjective experience, unique appraisal patterns and action tendencies which also held that each of the
4 persuasion (Nabi, 1999) and emotions conceptualized as mediators (e.g., Russell, 2003). These concepts will be used to theoretically explain the nature of distinct emotions.
Four discrete emotions of political communication
In (political) psychology (Basinger, 2008) and political communication literature, emotions are usually conceived of as "internal mental states representing evaluative reactions to events, agents, or objects that vary in intensity […] They are generally short-lived, intense, and directed at some external stimuli" (Nabi, 2002, p. 289–290). Emotional states considered in this present study are feeling hope, anger, fear, and empathy towards immigrants on the labor market since they are substantively associated with the topic (Nabi, 1999), present in the news media and reflect the public discourse. Ordinary citizens react to policies about the integration of immigrants on the labor market with discrete emotions such as fear and anger that perceived “illegals” pose to the national economic well-being but also with empathy and hope that
immigrants will fill more than a million empty positions in Germany due to an aging population and a shrinking labor force (Nienaber, 2018).
From a research perspective, emotional responses to groups that threaten (physical) safety have been a significant topic of research in political communication (Marcus, 2000). Fear and anger have attracted the most attention in the extant literature (Nabi, 2003) regarding their distinctive functions and action tendencies as well as being aroused in different social contexts such as immigration. To broaden the evidence of positive emotions, hope will be included because it is of interest when measuring the persuasiveness of news media (Myers, Nisbet, Maibach, & Leiserowitz, 2012). The influence of empathy is not investigated much in political communication but could be necessary for facing immigration (Lecheler et al., 2015) because it might help to understand other’s perspectives. These four discrete emotions are the first building block of this study. They will be contrasted due to their conceptual difference and differing effects as strongly emphasized in psychological and political literature (e.g., Lerner & Keltner,
5 2001) – even if they seem similar at first sight.
Appraisal theories of emotion
Recent scholarship explains the interplay of cognition and emotion within political
communication mainly with one dominant theoretical framework: appraisal theories. Invented by Magda Arnold in the 1960s who first used the term appraisal as an explanation of the elicitation of differentiated emotions, appraisal theories became essential when emotions are studied in experiments (Moors, Ellsworth, Scherer, & Frijda, 2013). In this purpose, the appraisal of a context-specific situation is towards the new immigration law in Germany. This, in turn, evokes an emotional response (or even no emotions) and can lead subsequently to an action (Frijda, 1986), e.g., behavioral intentions. The theory posits that individuals experience emotions as a product of their cognitive evaluation of a given event. This is the reason why emotional arousal might be individually different in similar situations (Lazarus, 1991; Scherer, 1999). Further, the underlying psychological concept of this investigation will refer to the
classical role of emotions in political communication: emotions as triggers. The basic idea is that emotions function as mediators between environmental stimulation and attitudinal as well as a behavioral response (Ellsworth & Smith, 1988b; Roseman, 1991). Based on these theories a growing body of literature has investigated the role of emotions in political communication, for example, the impact of emotions on information processing (Brader, Marcus, & Miller, 2011; Valentino, Gregorowicz, & Groenendyk, 2009) legal decision making (Nuñez, Schweitzer, Chai, & Myers, 2015) political judgments based on trust (Otto, 2018), and political participation (Valentino, Brader, Groenendyk, Gregorowicz, & Hutchings, 2011). Most relevant for this purpose are studies on opinion formation (Banks, 2014), on social judgment (Lerner & Keltner, 2001), policy support (Goodall, Slater, & Myers, 2013), attitudes (Nabi, 1999), and political behavior (Redlawsk & Lau, 2012).
6 Appraisal Dimensions
Appraisal theorists (Roseman, Spindel, & Jose, 1990; Scherer, 1999; Siemer, Mauss, & Gross, 2007) suggest a consensus that emotions are conceptualized by appraisal patterns, which are the critical concept of emotion-specific influences on judgment and choice. I borrow this idea from psychology and transfer it to political communication since it has not been included and thoroughly examined yet. For the first time in this research field, appraisal patterns will be manipulated to make predictions concerning the impact of distinct emotions.
Although a few theorists, most notably Roseman (1984) and Scherer (1982), have
independently proposed dimensional cognitive appraisal models, I mainly draw on Smith and Ellsworth (1985) because they present a summarized framework which usefully differentiates emotional experience and effects. They conclude that the experience of emotions is
conceptualized along six cognitive themes (appraisals). The primary appraisal is how pleasant the interpreted situation was (pleasantness). This dimension is also often called valence (positive vs. negative) of emotion and refers to earlier work that explains the impact of emotions on social judgments with the so-called valence-models (Watson & Clark, 1992; Watson, Wiese, Vaidya, & Tellegen, 1999). Consequently, all positive emotions were thought to positively influence dependent variables, while negative emotions reveal an association merely with adverse effects (Scherer & Moors, 2019). However, this does not fully explain the different effects of discrete emotions as shown in the studies above. Secondary appraisals differentiate emotions further and determine when the interpretation of the environment elicits explicitly emotions (Banks, 2014). For instance, how much the situation was out of anyone’s control (situational control), or someone who is responsible for the situation (human control) (Ellsworth & Smith, 1988a). Third, certainty is split into concerning the eliciting event and what would happen next, and the predictability of its consequences (Reisenzein & Spielhofer, 1994). Smith and Ellsworth (1985)
7 proposed further appraisal dimensions such as anticipated effort4, self-other agency5, and
attentional activity6, but all the investigated emotions will be associated with high levels of
anticipated effort and barely differ in responsibility and novelty. Thus, they do not offer distinction and will therefore not be part of the study. Instead, pleasantness, certainty and situational or human control are considered to be decisive factors for attitudes and behavior. In this study, these appraisals define whether emotional reaction to a newspaper article is attributed to costs (negative valence) or benefits (positive valence) of immigrants on the labor market, uncertain or certain consequences of the new law and human control (in this case the German government with chancellor Merkel) or situational control (the German economy).
Additionally, Ellsworth and Smith (1985) indicated that not all appraisals are equally relevant for all emotions and subsequent effects; especially certainty and control are two of the most often discussed appraisal dimensions over a threatening stimulus (Valentino et al., 2009). Further, pleasantness, situational control, and certainty are the prototypical appraisals for empathy (Wondra & Ellsworth, 2015). Finally, it is important to note that some emotions display overlapping patterns of appraisals such as fear and hope for which uncertainty is
paramount (Smith & Ellsworth, 1987). Emotions of the same valence such as anger and fear can differ in multiple appraisal patterns, e.g., concerning control and certainty (Lerner & Keltner, 2001). Nevertheless, both anger and fear are based on an unpleasant appraisal which indicates that respondents might feel more than one emotion, arguing for the arousal of blended or mixed emotions (Scherer, 2018).
I build my research upon Smith and Ellsworth (1987) who examined the relations between appraisals and emotions in the context of an upcoming midterm exam and the receipt of the
4 how much effort was required to cope with the situation for example, in fight or flight situations or on the other
side of the effort spectrum to relax, enjoy or withdraw quietly.
5 to what extent oneself, someone or something else was responsible for the situation.
6 how much we attend to a stimulus/situation and thus ignore or avoid it (attentional activity), which is similar to
8 exam grades. For instance, they found that anger was conceptualized by unpleasantness, while certainty was less critical. Ellsworth and Smith (1988a, 1988b) manipulated appraisals along the dimensions of high effort and uncertainty about the situation to test emotional differences of pleasant and unpleasant experiences. Moreover, Nabi’s (2003) and Goodall and colleagues’ findings are conceptually crucial as they suggest and manipulate content differences (absence or presence of alcohol) – which will also be the case in this present study – to investigate emotional reactions to the news on policy support. Balzarotti and Ciceri (2014) showed that the arousal of fear in the case of traumatic events was mediated by the appraisal dimensions of unpleasantness and coping potential. In line with this research, I provide further evidence for a causal link between appraisals and their corresponding emotions.
Which emotion plays what role?
To combine the theoretical assumptions of appraisal patterns and discrete emotions, I aim to answer the question of the characteristics of each investigated emotion.
Anger is an unpleasant emotion. According to appraisal literature, anger may evoke when an
individual is threatened and/or experiences demeaning offenses against oneself (Smith & Kirby, 2004). Anger is typically associated with the perceived certainty over the likely outcome of a situation (Roseman, 1984). Moreover, anger is elicited by a perceived human control, i.e., a person who is blameworthy and responsible for the offense (Banks, 2014).
Fear is a negatively-valenced emotion, and cognitive appraisal theories argue that fear
results from the appraisal of threat and danger (Witte, 1992). It occurs when people are
uncertain about what will happen (Tiedens & Linton, 2001). These unknown and uncontrollable events may evoke appraisals of uncertainty (Slovic, 1987). Situational control leads to high rates of fear (Frijda, 1986).
9 nature of human emotions (and also the researcher's theoretical biases (Ellsworth & Smith, 1988a)), this present study attempts to map out the terrain of positive feeling states.
Frederickson (1998, 2001) argued (but did not test) that emotions with similar (positive) valance may differ in emotion-specific effect on behavior similar to negative emotions. However, little is known about it (Lecheler, Schuck, & de Vreese, 2013). Nussbaum (2013) argues that in order to persuade citizens, powerful positive emotions should be activated because they can motivate and sustain citizens' cooperation which is needed for constructing a better society and the integration of immigrants.
Hope is associated with “warmth-friendliness” (Roseman, 1991, p. 161) and thus, a pleasant
emotional experience. Further, uncertainty is one of the main appraisal patterns of hope
(Ellsworth & Smith, 1988b). People feel least certain in a hopeful situation when an unexpected situation with a positive outcome happens. Further, hope is highly consistent with situational control.
Empathy7, agreed upon by most empathy researches (e.g., Bartsch, Vorderer, Mangold, &
Viehoff, 2008; Wondra & Ellsworth, 2015) is the feeling another person feels which makes it difficult to categorize as pleasant or unpleasant. The emotional experience of empathy is regarded as an unusual phenomenon because it differs from usual firsthand experienced
emotions. Sometimes we feel emotions because something happens to someone else. Those are called vicarious emotions (Lamm, Batson, & Decety, 2007). It might be a relevant concept to consider in political communication since we often experience emotions of a political event or a policy vicariously rather than directly related to us. Wondra and Ellsworth (2015) proposed that prototypical empathy would include appraisals of high or low pleasantness, moderate certainty, and upper situational control. Further, empathy is associated with the ability to take the other
7 The most comprehensive view of empathy in psychological research through the 20th century is provided by the
moral development theory (Hoffman, 2000). Brain research is the dominant approach since the 2000s, such as mirror neuron (Rizzolatti, Fadiga, Gallese, & Fogassi, 1996) and perception-action theories (Preston & de Waal, 2002)however, they are less remarkable when applied to empathy.
10 person's perspective and could lead to caring for others (Decety, 2011; Zaki, 2014) and to
altruistic behavior (Toi & Batson, 1982). Based on the positive outcomes of empathy, I consider it as a somewhat positive emotion than a negative one, although it is regarding the valence not necessarily either the one or the other.
Table 1
Appraisal patterns of emotions
Anger Fear Hope Empathy
Appraisal pattern
Pleasantness Low Low High High
Certainty High Low Low Moderate
Control Human Situational Situational Situational
Note. Theoretical assumptions (Smith and Ellsworth, 1985; Wondra and Ellsworth, 2015). Since a direct link between appraisal categories and response patterns exists (Lazarus, 1991; Roseman, 1991; Scherer, 1999; Smith & Ellsworth, 1985), contrasting these four emotions seems reasonable to explore the idea of appraisals eliciting emotions. Therefore, I hypothesize: Hypotheses block I – Appraisals and emotional arousal
H1a: The combination of unpleasant information, high certainty, and human control appraisal in a newspaper article will elicit more anger than any other condition.
H1b: The combination of unpleasant information, low certainty, and situational control appraisal in a newspaper article will elicit more fear than any other condition.
H1c: The combination of pleasant information, low certainty, and situational control appraisal in a newspaper article will elicit more hope than any other condition.
H1d: The combination of pleasant information, moderate certainty, and situational control appraisal in a newspaper article will elicit more empathy than any other condition.
11 Appraisal Tendencies and Attitudes
The importance of studying appraisals can be connected to the capacity of emotions to affect support of an issue (Goodall et al., 2013) as well as attitudes (Nabi, 1999). Overcoming the limitations of the valence models to emotions, Lerner and Keltner (2000) call the appraisal dimensions above appraisal tendencies. Based on the common assumption that cognitive appraisals cause emotions, appraisal tendencies predispose individuals to appraise the situation in line with goal-directed processes through which emotions exert effects on attitudes. As attitudes are also built on cognitive assumptions (Eagly & Chaiken, 1993), cognitive effects have a significant impact on supporting or rejecting an issue: emotional congruent cognitions (=appraisals) promote emotionally congruent attitudes. Other approaches postulate that emotions are used as judgment heuristics (Forgas, 1995), i.e., emotions can be projected on objects, in this case, the new immigration law and thereby influence the rating of the object. For instance, anger at immigrants might result in negative policy support. It should be noted that attitudes elicited by emotions are more profound than cognitively formed attitudes (Kühne, 2013).
Of course, there are other variables such as political efficacy (Valentino et al., 2009), predisposition (MacKuen, Wolak, Keele, & Marcus, 2010), partisan bias (Weeks, 2015) that influence emotional experience regarding immigration. However, this study does not focus on citizens’ angry or hopeful reaction when being exposed to news media but rather, to what extent appraisals evoke emotions and subsequently influence attitudes.
Prior research suggests the possibility that appraisals explain discrete emotional reactions to a particular topic in the news and their effects on policy support/rejection as well as attitudes. Although Lecheler et al. (2015) test framing effects about immigration mediated by positive and negative emotions, they assume that each news frame causes different topic-relevant emotions based on appraisals. Nabi (2003) and Huddy, Feldman, Taber, and Lahav (2005) showed that fear and anger (two emotions with the same negative valence) resulted in different effects; while
12 fear promoted more protective policies, anger supported more punitive policies. Similarly, Goodall et al. (2013) demonstrated that existing laws according to alcohol-related crimes are enforced when anger predominated, whereas new regulations are supported when fear dominated. Based on these findings, this study also assumes differences in the attitudinal outcomes of fear and anger.
Most relevant for this study, hope was increased in response to a positive yet uncertain outcome (Roseman, 1991). As one of the few studies exploring positive emotions, Lecheler et al. (2015) found different mediation patterns of immigration frames demonstrating that the emancipation and assimilation frame were mediated by hope. Further, hope may play a role in affecting political attitudes (e.g., Brader, 2005; 2006) and has been found to positively affect opinion formation about immigration (e.g., Griskevicius, Shiota, & Neufeld, 2010). For instance, feelings of hope drive candidate evaluations and exchange appraisals for a better alternative (Just, Crigler, & Belt, 2007). In addition, empathy leads to immigration policy
support (Verkuyten, 2004). In line with this result, other specific positive sentiments such as pity and sympathy for a woman caused by an episodic frame were linked to increased oppositional feelings to mandatory minimum sentencing (Gross, 2008).
At heart of this study is to investigate whether and how the role of anger, fear, hope, and empathy will affect policy support or policy rejection, as well as attitudes towards immigration on the German labor market, are defined. As anger is elicited by negative valence, certainty about the situation as well as human control attributed to a clear agent with the German
government and Chancellor Merkel, I expect that anger will reject the policy and correlates with the negative attitude towards immigration on the labor market. Evidence was found for angry voters when they were able to blame someone responsible for a negative perceived political event, for example, a government official (Valentino et al., 2011). Even though fear should be affected by the fearful stimulus, I expect that the emotion of fear will be uncorrelated to the
13 policy and attitudes due to uncertainty about the situation and vague situational control rather than an explicit control in the fear condition (Lerner & Keltner, 2000, 2001; Otto, 2018). Hope is also conceptualized by uncertainty and situational control but shedding a positive light on the policy and keeping prior findings in mind; I expect that hope will be correlated with policy support and positive attitudes towards immigration. As a reader feels somewhat confident that the refugee could stay because of the new law (Wondra & Ellsworth, 2015), I hypothesize empathy will be correlated with policy support and positive attitudes towards immigration on the German labor market.
Hypotheses block II – Policy support and rejection
H2a: Anger mediates the effect of news exposure on policy rejection.
H2b: Fear is uncorrelated to the effects of news exposure on the new immigration law. H2c: Hope and empathy mediate the effect of news exposure on policy support.
Hypotheses block III – Attitude towards immigration on the labor market
H3a: Anger mediates the effect of news exposure on a negative attitude towards immigration on the labor market.
H3b: Fear is uncorrelated to the effects of news exposure on attitude towards immigration on the labor market.
H3c: Hope and empathy mediate the effect of news exposure on a positive attitude towards immigration on the labor market.
Behavioral intentions
Functionalist approaches to emotion draw on the same basic concepts as appraisal theories but within the emotional experience, one fundamental principle is of particular importance: each emotion serves a distinctive goal or motivation represented in action tendencies and are
therefore central in the analysis of emotions as such (Frijda, 1988; Nabi, 1999; Ortony, Clore, & Collins, 1988; Scherer, 1984). Based on the assumption that action tendencies influence
14 behavior, behavioral intentions can be taken as an indirect measure of action tendencies
(Martinez, Zeelenberg, & Rijsman, 2011). Behavioral intentions describe the probability a person will perform a particular behavior in the future (Ajzen & Fishbein, 1980).
A long line of research acknowledged that emotions are arguably the primary motivation for human behavior (Izard, 1977). For instance, negative emotions can break people out of political habits (Brader, 2005, 2006). Marcus and colleagues' findings (2002) hold important implications for political behavior since anger is typically associated with action out of the threat. Anger is also an approach emotion that motivates to strike out and is closely tied to approach behavior (in the form of lashing out) (Goodall et al., 2013), for instance discussing a topic with family and friends, taking part in demonstrations, or signing a petition. In contrast, fear is often equated with avoidance and withdrawal tendencies (Frijda, Kuipers, & ter Schure, 1989; Smith, Haynes, Lazarus, & Pope, 1993). It activates energy to escape from the threatening agent (Newhagen, 1998; Roseman, 1984). From a functional perspective, positive emotions encourage approach behavior towards rewarding stimuli (Huddy, Feldman, & Cassese, 2007). Hope is highly consistent with the function of overcoming perceived goal obstacles and approach behavior (Roseman, 1991; Smith & Ellsworth, 1985). These findings again confirm that it is essential to distinguish between anger, fear, hope, and empathy to understand their distinct effects. Based on this evidence, I predict the following:
Hypotheses block IV – Behavioral intentions
H4a: Anger mediates the effect of news exposure on approach behavior.
H4b: Fear mediates the effects of news exposure on withdrawal and avoidance of political action.
H4c: Hope mediates the effect of news exposure on approach behavior.
To broaden the evidence of a mediating effect of positive emotions, I include empathy as a positive emotional state, which has not been thoroughly examined in political communication
15 but seems important in the context of immigration (Lecheler et al., 2015). Thus, I do not have an explicit assumption and pose the following research question:
RQ1: What is the role of empathy between the effects of news exposure on behavioral intention and how will empathy influence behavioral intentions?
Figure 1. Variable Model
Method General Design and Procedure
I used a 5 x 1 pre- and post-test experimental design with one between-subject factor representing the five conditions (anger, fear, hope, empathy, control). The survey embedded experiment was conducted from 15th to 19th December 2018 within AScoR to analyze appraisal effects on emotions and subsequent implications on attitudes and behavioral intentions. As this study does not follow a standard procedure, I conducted a pilot study from 10th to 13th of December 2018 with a convenience sample to pretest the stimuli, the appraisal dimensions, and manipulation checks (see Appendix A). It revealed that the empathy and anger condition elicited significantly more empathy and anger than any other condition. However, neither the fear nor the hope condition elicited significantly more fear or hope in comparison to the other conditions. As both emotions fear and hope are characterized by uncertainty, I made the appraisal
Policy support/rejection, attitudes, behavioral
intentions Anger, fear, hope,
empathy
16 manipulation in the newspaper articles more visible. Due to the unsuccessful manipulation checks of certainty and control, I adjusted the questions and improved them in the main study. The data for the main study was collected with Mturk and Clickworker in Germany, two
crowdsourcing platforms which enable survey distribution. Participants were randomly assigned to one of the four conditions - hope, anger, fear, empathy - or a control group through block randomization.
Stimuli
The strength of emotional elicitation which is triggered by content manipulation of
appraisals depends on features of the text such as the thematic relevance, the structure, and the degree of realism of the presentation (Ortony et al., 1988). Using an actual hot topic also
enhances the realism of the experiment (Myers et al., 2012) since both the negative and positive consequences of the new immigration law is to be found in real political news coverage. I undertook significant effort to design the articles in a conventional layout and editorial style of German newspaper articles using information from Süddeutsche Zeitung, Frankfurter
Allgemeine Zeitung, Die Zeit, and Bild. I thereby endeavor to provide an as close to a "real world" test of appraisal dimension to elicit emotions as is realistically predictable in an experiment.
For each of the four experimental news stories (see Appendix B/C for German and English), I chose one emotion – hope, anger, empathy, and fear – relevant to the context. Following the example of other immigration studies (see, e.g., Igartua et al., 2011; Lecheler et al., 2015; Verkuyten, 2004), the four articles contained 541 to 549 words. The headline, introduction as well as the final paragraph were kept identical in all four conditions. The control group was only composed of these three parts (124 words). The introduction offered specific information about the law. The final paragraph generally alluded to the expectation that the new ruling is passed by
17 the end of 2018. The core section of each experimental news story contained a direct appraisal manipulation by presenting real events which I expect to be evaluated in a certain way (Scherer & Moors, 2019) (see Appendix D for a tabular overview of appraisal manipulation and news content).
(Un-)Pleasantness. The hope and empathy conditions presented the law in a positive way
referring to three positive consequences such as rise of economic productivity, lower prices of goods and services, increase of general living standard, while the anger and fear conditions displayed negative consequences such as lower wages, preferential treatment of refugees, and encouragement of further immigration (see Appendix E).
(Un-)Certainty. This core section was framed by two sentences at the beginning and two
sentences at the end about the extent of (un-)certainty about the consequences of the new law. For instance, the anger condition pointed out that the government is willing to improve on detail questions, and the predicted consequences will be embedded in the new law. To calculate the implications of the law, the government will take the particular action. In comparison, fear and hope presented very vaguely, unclear impacts of the law. The reader should be uncertain about how to deal with the situation and feel unable to cope. In the empathy condition, the reader should feel somewhat confident about the situation.
Human or situational control. Whenever the anger condition offered the possibility to explain who introduced the new law, the German Government with Angela Merkel (=human control) was mentioned. Based on standard measurement of anger (see Harmon-Jones et al., 2016) it should create a sentiment of blaming someone (the German government) for something terrible (the law) because the new regulation is allegedly Merkel’s fault which harms, for instance, long-term unemployed and prevents them from getting something they want (=jobs). Situational control for the hope, fear and empathy condition was mainly expressed with the German economy initiating the new law and barely referring to the German government. In the
18 empathy condition, Abel is grateful for Germany and the opportunities the country has given him.
Measurement
Emotional reaction. Respondents were asked to what extent they felt the emotions hope, anger, empathy, fear while reading the newspaper article on an seven-point Likert-scale (1-not at all; 7-very much) based on the differential affect scale (Renaud & Unz, 2006). As the fear and anxiety item correlate highly (a = 0.88), I combined the two measures into a composite fear scale.
Policy support/rejection. This first dependent variable was operationalized using the prompt “To what extent do you agree with the new law or do you reject the new law?” by applying a seven-point Likert-scale (1-strongly reject; 7-strongly agree) (e.g., Goodall et al., 2013).
Attitudes. The second dependent variable was operationalized with the question “To what extent do you agree with the following statements?” on a seven-point Likert scale (1-not at all; 7-very much) (e.g., Druckman & Nelson, 2003). Ten items asked about whether immigrants enrich German culture, are a threat to the labor market or misuse social welfare8. The items did
not correlate strongly (a = 0.31), but a principal component analysis (PCA) revealed five items correlate positively with the first component (eigenvalue 4.68), which appears to measure costs of immigration on the German labor market (a = 0.83) (M = 20.21, SD = 6.82). The other five items correlated positively with the second component (eigenvalue 1.44) which defines the benefits of immigration on the German labor market (a = 0.82) (M = 23.06, SD = 5.76).
8 More items: (4) The German government targets immigrants as a priority group. (5) The German government has
to protect immigrants from hostility towards foreigners. (6) Hostility towards foreigner is harmful to the economy. (7) Due to the increased number of immigrants, I sometimes feel like a foreigner. (8) Criminal immigrants should be more consistently deported to their countries of origin. (9) Immigration offers the opportunity to mitigate the skilled labor shortage in Germany. (10) Immigrants take on jobs that Germans do not want to take over.
19 Behavioral intentions. The third dependent variable was measured with seven items (a = 0.80) such as the probability of talking about the immigration law with family and friends, take part on a demonstration for migration, sign up a petition, look for more information about the topic of migration on a seven-point Likert-scale (1-highly unlikely; 7-very likely) (Marcus, Neuman, & MacKuen, 2000)9. Higher values indicate approach behavior (anger and hope),
while lower values refer to avoidance behavior (fear) (Huddy, Feldman, Taber, & Lahav, 2005).
Participants
The sample was originally composed of 434 participants; however, 35 participants did not complete the survey and were therefore excluded from the data set. The final sample (N = 399) had a greater proportion of males (57.6% male, 41.1% female10), was rather young (M = 35.69,
SD = 12.33; 16 to 71 years) and diverse. 17% of the participants had a migration background11.
The political orientation was slightly right wing (M = 5.51, SD = 1.99; 0-left-wing; 10-right-wing) and well-educated (45.6% of the participants had a university degree). Though the sample was not representative for the German electorate, the participants' background displayed good variation. The randomization check revealed successful randomization with no between-group differences for the sample on age and education but not on gender (see Appendix F). There were significant differences between women and men in the hope and control condition
(conditionhope: 29.2% female, 70.8% male; conditioncontrol: 54.5% female, 45.4% male; χ2 (5,
399) = 1.461, p = 0.057), which means that gender needs to be included as a covariate in the analysis.
9 More items: (5) Get in touch with immigrants. (6) Work together with immigrants. (7) Donate money for a
charitable organization taking care of immigrants.
10 1.3% of the participants did not specify their gender.
20 Manipulation Checks
Pleasantness
The manipulation check question for pleasantness asked respondent whether the article they had just read shed a positive or a negative light on the new immigration law (1 = negative; 5 = positive) (e.g., Lecheler et al., 2013) and revealed successful manipulation. Participants rated the hope and empathy conditions significantly higher than the anger and fear condition (Mhope = 3.75, SDhope = 0.77; Manger = 2.66, SDanger = 0.80; Mfear = 2.83, SDfear = 0.69; Mempathy = 3.99, SDempathy = 0.93; F(5,393) = 34.24, p = .000). The post-hoc-tests (Bonferroni) revealed that there is no significant difference between the hope and empathy condition (M = -.23, SD = .14, p > .05), nor for the fear and anger fear condition (M = .17, SD = .15, p > .05).
Certainty
The manipulation check for certainty was composed of a semantic differential (a = 0.88). Six item pairs (e.g., vague-clear, uncertain-certain, unpredictable-predictable) were derived from the Geneve Appraisal Questionnaire (Scherer, 2001). Participants were asked how they would evaluate the consequences of the law on a seven-point scale. Lower values indicated uncertainty, while higher values referred to certainty. The manipulation check yields significant differences for the evaluation of certainty: Participants in the anger condition perceived the consequences of the law as more certain than in the other conditions (Mhope = 4.35, SDhope = 1.18; Manger = 4.60, SDanger = 0.99; Mfear = 3.77, SDfear = 1.17; Mempathy = 4.49, SDempathy = 1.31; F(5,393) = 4.34, p = .001). The Bonferroni post-hoc test revealed a significant difference between anger and fear (M = .83, SD = .21, p < .01), and fear and empathy (M = -.71, SD = .20, p < .01). However, there should also be a significant difference between the anger and hope (M = .25, SD = .20, p > .05)
condition since they theoretically differ in certainty. Thus, the manipulation check was partly successful.
21 Situational and human control
As the pilot study revealed that situational and human control is difficult to measure, I used an open answer category asking participants who is responsible for the law whether is the
German government (human control) or the German economy (situational control). Each answer was coded in either 1=government, 2=economy, 3=both, 4=non of them, or 5=other (Bryman, 2016). Categories 3, 4, 5 were excluded because the manipulation of control is either provided by the government or the economy. The results indeed show successful manipulation.
Participants in the anger condition scored lower than the other conditions which confirms that the human control appraisal was more perceived in the anger condition. (Mhope = 1.79, SDhope =
0.41; Manger = 1.38, SDanger = 0.49; Mfear = 1.66, SDfear = 0.48; Mempathy = 1.73, SDempathy = 0.43; F(5,260) = 4.667, p = .000). However, the test of homogeneity is significant, which does not meet the criteria for an one-way ANOVA. The Bonferroni post-hoc test revealed a significant difference between anger and hope (M = -.41, SD = .10, p < .01), and anger and empathy (M = -.35, SD = .10, p < .05), however, there is no significant difference between anger and fear (M = -.27, SD = .09, p > .05). Thus, the manipulation check was partly successful.
Results
The purpose of this study is first to test whether appraisals evoke emotional states (hypotheses block I), and second whether emotional responses to media stimuli mediate attitudes and behavioral intentions towards immigration on the German labor market
(hypotheses block II – IV). I apply a method by Hayes (2012) to test for mediation as a causal approach. Therefore, the multicategorical independent variable was dummy coded with five experimental groups 0 = control group and reference category, 1 = hope, 2 = anger, 3 = fear, 4 = empathy (Hayes & Preacher, 2014).
22 Hypotheses block I – Appraisals and emotional arousal
I first conducted a one-way ANOVA to check whether each condition was successful in generating the corresponding emotions. Both the fear condition F(5,392) = 3.587, p < 0.01 and the empathy condition F(5,393) = 5.270, p < 0.001 elicited significantly more fear and empathy than any other condition, as expected in H1b and H1d. In contrast, the hope condition F(5,393) = 8.492, p < 0.001 could not arouse more hope than the empathy condition, whereas the anger condition F(5,392) = 3.775, p < 0.01 did not generate significantly higher amounts of anger than the fear condition12. Thus, H1a and H1c have to be rejected.
Table 2
Means and standard deviations for emotions after reading the five conditions Conditions
Baseline Hope Empathy Anger Fear
M SD M SD M SD M SD M SD Emotions Hope 3.91 1.79 3.93 1.44 3.97 1.44 3.27 1.72 2.69 1.44 Empathy 3.19 1.61 3.34 1.72 4.37 1.51 3.37 1.79 3.52 1.65 Anger 2.37 1.67 2.70 1.52 2.66 1.56 3.27 1.68 3.42 1.64 Fear 2.28 1.61 2.34 1.63 2.32 1.61 2.73 1.99 3.16 1.81 Note. N = 399.
Hypotheses block II – Policy rejection or support
I conducted mediation analyses to investigate whether emotional responses to media stimuli mediate the subsequent policy support or rejection. Empathy, fear, and anger are affected by the newspaper articles with the appraisal manipulation (see a-path coefficients in figure 2), and also
12 The Bonferroni test revealed that there are significant differences between the fear condition and control group
(M = .88, SD = .27, p = .015), and the empathy and the control group (M = 1.18, SD = .25, p = .000). Neither the hope (M = -.02, SD = .28, p = 1.000) nor the anger condition (M = .89, SD = .31, p = .055) are significantly different from the control group.
23 show significant indirect effects of anger on policy rejection, thus confirming hypothesis H2a (see table 3). The direct effect for anger rendered nonsignificant for the appraisals in the newspaper articles suggesting a full mediation. I expected fear to be uncorrelated to the new policy; however, fear can explain variance on policy rejection above and beyond gender (see Appendix F). Thus, H2b has to be rejected.
There is no significant mediation effect of hope, although the emotion hope leads to policy support (see b-path coefficients in figure 2)13.
Moreover, the newspaper article with the appraisals for empathy leads to elicitation of empathy. Nevertheless, the results show no significant mediation effect by empathy regarding policy attitudes. Consequently, H2c needs to be rejected. The overall mediation model can be used to predict policy support or rejection towards the law F(10,382) = 12.32, p < .001 but the strength of the prediction is moderate (R2 = .24).
Figure 2. Mediation model of indirect effects of newspaper articles with appraisals on attitude about the law through discrete emotions. All coefficients are unstandardized. *** p < .001, **p < .01.
13 In contrast to that, the fear condition is negatively correlated with the elicitation of hope (b = -1.15, SE = .28, p <
.001).
Appraisals Attitude towards
the law Hope Empathy Fear Anger .10 1.26*** .92** .92**
No significant direct effects
.40*** .11
-.35*** -.30**
24 Table 3
Indirect effects on policy attitudes
95% CI IND SE LL UL Mediated via Hope .407 .089 .231 .583 Empathy .105 .098 -.087 .297 Anger -.349 .088 -.523 -.172 Fear -.308 .095 -.495 -.120 Gender -.308 .247 -.794 .176
Note. N = 399. CI = confidence interval; LL = lower limit; UL = upper limit.
Hypotheses block III – Attitude towards immigration on the German labor market
According to the low correlation between the original ten items, the scale had to be split into two variables – cost and benefits of immigration on the labor market. Both mediation models can be used to predict attitudes towards costs on immigration F(10,380) = 7.99 , p < .001 with a low strength of prediction (R2 = .17) and attitudes towards benefits on immigration F(10,380) = 11.87, p < .001 with a moderate strength of prediction (R2 = .24). Table 4 and 5 show a
significant mediation through anger (confirming H3a) and fear (rejecting H3b), which are positively related to attitudes toward costs and negatively related toward benefits of
immigration. Fear can explain variance above and beyond gender. The direct effect for fear rendered nonsignificant for the appraisals in the newspaper articles suggesting a full mediation. The newspaper article with the anger appraisals has a direct effect on the benefits of
immigration (see figure 4). Although exposure to the newspaper article with hope appraisals did not cause a feeling of hope, I can show a significant positive relationship of hope on the benefits and a negative effect on costs of immigration. Empathy served indeed as a mediator on attitudes towards benefits of immigration; however, it did not mediate the relationship between the effects appraisals have on the costs of immigration. Even though these results partly confirm my theoretical assumptions, H3c has to be rejected.
25 Figure 3. Mediation model of indirect effects of newspaper articles with appraisals on attitude on costs of immigration through emotions. All coefficients are unstandardized. *** p < .001, **p < .01, *p < .05.
Table 5
Indirect effects on attitudes towards costs of immigration
Boot CI IND SE LL95 UL95 Mediated via Hope -.111 .035 -.179 -.043 Empathy -.053 .038 -.128 .021 Anger .070 .034 .003 .137 Fear .137 .037 .064 .209 Gender .075 .096 -.114 .264
Note. N = 399. CI = confidence interval; LL = lower limit; UL = upper limit.
Appraisals Attitude on costs
of immigration Hope Empathy Fear Anger .12 1.29*** .91** .90** -.11** -.05 .07* .14*** No significant direct effects
26 Figure 4. Mediation model of indirect effects of newspaper articles with appraisals on attitude on benefits of immigration through emotions. All coefficients are unstandardized. ***p < .001, **p < .01, *p <.05.
Table 6
Indirect effects on attitudes towards benefits of immigration
95% CI IND SE LL UL Mediated via Hope .123 .034 .057 .189 Empathy .121 .037 .048 .193 Anger -.112 .033 -.177 -.047 Fear -.105 .036 -.175 -.035 Gender -.080 .093 -.262 .103
Note. N = 399. CI = confidence interval; LL = lower limit; UL = upper limit.
Hypotheses block IV - Behavioral intentions towards immigration
Empathy fully mediated the effect of the newspaper article with empathy appraisals on behavioral intentions (RQ1). The emotion significantly increased the probability that
participants in the empathy condition will talk to family and friends about immigration, take part in demonstrations or sign a petition on immigration on the labor market. The direct effect for empathy rendered nonsignificant for the appraisals in the newspaper articles suggesting a full
Appraisals Attitude on benefits
of immigration Hope Empathy Fear Anger .13 1.29*** .92** .91** Anger: .35* .12*** .12*** -.11*** -.11*
27 mediation. The mediation models of anger and fear were nonsignificant (see table 6). Although the hope treatment did not cause hope, it had a positive effect on behavioral intentions.
Nevertheless, H3a-H3c have to be rejected. The overall mediation model can be used to predict behavioral intentions F(10,356) = 7.30 , p < .001 but the strength of the prediction is weak (R2 = .17)
Figure 5. Mediation model of indirect effects of newspaper articles with appraisals on behavioral intentions through emotions. All coefficients are unstandardized. ***p < .001, *p <.05.
Table 7
Indirect effects on behavioral intentions
95% CI IND SE LL UL Mediated via Hope .150 .030 .092 .209 Empathy .080 .033 .016 .145 Anger .039 .029 -.019 .096 Fear .001 .032 -.062 .065 Gender -.014 .081 -.174 .145
Note. N = 399. CI = confidence interval; LL = lower limit; UL = upper limit.
Appraisals Behavioral intentions Hope Empathy Fear Anger .09 1.21*** .83* .84* .15*** .08* .04 .01 No significant direct effects
28 Discussion
Emotions are considered integral to understanding political communication (Groenendyk, 2011). The findings show the importance of appraisals and their ability to arouse the
corresponding emotion especially in the case of empathy and fear condition (hypotheses block I). Conducting appraisals along dimensions that discriminate among responses, the results indicate that pleasantness seems to be critical for triggering emotion. Certainty is likely to be relevant for the arousal of anger, moderate certainty for empathy, and uncertainty for fear. Hope and empathy seem to be characterized by situational control, while anger is conceptualized of human control. Moreover, my findings consistently support that emotions can explain variance on attitudes and behavioral intentions towards immigration on the labor market in Germany (hypotheses block II-IV). Anger and fear mediate the effects of appraisals on the rejection of the new immigration law and negative attitudes towards immigration on the labor market. Further, this study illustrates that hope has the highest impact on supporting the new immigration law, positive attitudes towards immigration and behavioral intentions, while empathy mediates the effects of appraisals on behavioral intentions towards immigration.
Transferring these ideas and findings from emotions research in psychology to political communication, this paper provides four main contributions to the existing research:
Firstly, this paper is among the first who manipulated the frequently stated but rarely tested appraisal dimensions with varied content in newspaper articles. Examining a highly relevant and emotional induced issue, namely the new immigration law in Germany, this study does not only provide high external validity but also demonstrates that appraisals are crucial predictors for emotional arousal in political communciation. This adds to the distinction of discrete emotions along specific appraisal patterns which goes beyond the usually tested valence and is consistent with Goodall et al. (2013), Nabi (2003), Smith and Ellsworth (1985).
29 Secondly, this study emphasizes the dominant role of negative emotions in immigration discourse drawing on the basic emotions anger and fear as a backbone of classic affect effect research (Gadarian & Albertson, 2014; Igartua et al., 2011; Lecheler et al., 2015). Although different appraisals elicited anger and fear, I could not pursue the notion of subsequent distinct effects (Brader, 2005, 2006) indicating the dominance of the primary appraisal of pleasantness (Scherer & Moors, 2019).
Thirdly, even though hope appraisals could not elicit more hope than any other condition, this study broadens the evidence for positive emotions and their effect on positive attitudes, which is in line with Griskevicius et al. (2010). As emotions are associated with action tendencies (Frijda, 1988), hope and empathy induce approach behavior. Strong positive
emotions seem to get citizens involved in politics (Groenendyk, 2011) corresponding to the idea of a functioning pluralistic, multicultural society. A positive presentation of a political issue (Boomgaarden & Vliegenthart, 2009), as well as positive emotions, could lead to caring within the larger society ensuring its stability over time (Nussbaum, 2013)
Fourthly, this study indicates that empathy should be given more consideration in political communication (Lecheler et al., 2015). Although it has not been investigated much so far, it seems to be the key to political persuasion, especially in the case of immigration and integration of refugees. Empathy allows us to understand different perspectives (Wondra & Ellsworth, 2015).
Despite the illustrated results and contributions, this study has a number of limitations and inconsistencies, which in some cases might lead to new ideas for future research. First, one experiment can only assess short-term effects on how media exposure (given that many stories may be encountered daily) influences citizens’ attitudes and behavioral intentions. Although external validity is high in this study, it is questionable how experimental findings can be conveyed to real-life media effects over time. My second point of discussion refers to the (non-) elicitation of emotions based on appraisals. For example, the hope condition did not show the
30 desired outcome; presumably because it is too similar to the control group. Another explanation could be the insufficient attention check (see Appendix G) and a low message perception (see Appendix H) for the hope condition. Further, the appraisal manipulation did not work
adequately which leads to my next limitation: the partly successful manipulation checks. Maybe it did not measure what it was supposed to measure or most likely appraisals cannot be
measured due to subconscious processes (Frijda, 1986). Scherer (2005) acknowledges that the measurement of emotions remains an important scientific task. In a similar vein, empathy as a vicarious emotions can cause problems when a person feels angry while another individual does not share the same state of emotion about an issue impacting a third person (Wondra, 2017). This is not a unique problem since the same terrible situation could make one person angry, while the other person feels nothing about it. Another emotion problem is that not necessarily only one emotion was elicited but multiple contributing to the concepts of mixed or blended emotions (Scherer, 2018) and the theory of emotional flow (Nabi, 2015). This new wave of research focuses on the effects of a collection of emotions instead of a singular emotional state, which might be relevant in the case of immigration debates since this topic is infused with several emotions. More emotional appeals such as pride, disgust, and contempt could be investigated in future research.
Showcasing the limited effects of emotions in general, it must be kept in mind that according to appraisal theories the evaluation of a situation is highly subjective and depends on how the individual appraises the situation (Lazarus, 1991). This paper cannot provide answers to these phenomena but assumes that the solution can be found in appraisals. Future studies should clarify these appraisal processes in more detail by investigating how appraisals interact, whether the order of appraisals matters and even discover new dimensions. Increased attention should be cast on the different targets of emotions based on the same emotional state.
31 Another limitation of this study is that emotions are conceptualized as mediators. From a theoretical point of view, it points out shortcomings since the appraisal theories suggest a dynamic and fast interplay between cognition and emotional responses. Emotions could also simply be a (side) effect of negative or positive evaluations of the new German immigration law. Following this from a methodological perspective, the causality assumption of these mediation analyses is problematic. I have to acknowledge that only the appraisals cause
emotions, while the relationship between the emotions and attitudes and behavioral intentions is correlational. While behavioral intentions by definition tend to follow attitudes (Ajzen &
Fishbein, 1980), the design does not empirically disprove that the participants’ attitude has an impact on certain emotions, nor can it rule out that it happens at the same time. Nevertheless, the results infer that emotions function as mediators because the conditions elicited the desired emotions and only some of them affected the subsequent attitudes and behavioral intentions, which is in line with the primacy of emotions affect hypothesis (Zajonc, 1980). This theory further asserts that the emotional response to the stimuli can be processed faster and more readily than their cognitive attribution in this case attitudes. In addition, the overall missing direct effects14 might indicate that there is no pure cognitive evaluation adding to the hot
cognition hypothesis (Abelson, 1963). It claims that objects related to politics are tagged with emotions. And indeed, immigration debates are highly emotional. Hence, emotions tend to influence attitudes rather than vice versa (Eagly & Chaiken, 1993).
Finally, emotions might be the missing piece in the jigsaw that scientist have struggled to find for the last decades. However, other pieces in a variety of puzzles exist such as political knowledge (Romano, 2018), need for affect (Maio & Esses, 2001), political efficacy (Valentino
14 Anger appraisals showed a direct effect on attitudes indicating that appraisals tend to influence the reader’s cognitive evaluation of an issue. Emotional reactions towards attitudes of immigration cannot be explained only by newspaper articles, but other factors must come into play.
32 et al., 2009), prior beliefs and preexisting attitudes (Banks, 2014). Yet, these moderating
variables were not taken into account in this study although they might be helpful to the issue of immigration. Future study should examine this in more detail.
Conclusion
This study is among the first who manipulated appraisals (and not frames) to broaden the evidence for underlying processes of media effects in political communication. Precisely, how news reports influence discrete emotions. From heads to hearts: appraisals and emotional appeals (not facts) guide us subconsciously. They affect our perception of the political world and consequently our attitudes towards immigration. Emotional rhetoric with positive and negative sentiments keep the immigrant issue on the political agenda. Thus, we should be aware of the fact that emotions tend to generalize. The rage and fear against immigrants on the German labor market felt by some participants could sustain the perception of immigrants as being generally bad and dangerous. Consequently, information about immigration must be reported carefully in the media, also to avoid stereotyping. Further, this study provides information on how citizens effectively convert that news - via emotions - into attitudes and express it through political participation. Especially positive emotions such as hope and empathy seem to motivate citizens to engage in politics. This finding raises hope for a better mutual understanding in society and an overall willingness to tolerate others’ behaviors for the good of democracy.
33 References
Ajzen, I. & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall.
Aarøe, L. (2011). Investigating Frame Strength: The Case of Episodic and Thematic Frames. Political Communication, 28(2), 207–226.
https://doi.org/10.1080/10584609.2011.568041
Abelson, R. (1963). Computer simulation of “hot” cognition. In S. S. Tomkins & S. Messick (Eds.), Computer simulations of personality (pp. 277–302). New York: Wiley.
Balzarotti, S., & Ciceri, M. R. (2014). News Reports of Catastrophes and Viewers’ Fear: Threat Appraisal of Positively Versus Negatively Framed Events. Media Psychology, 17(4), 357–377. https://doi.org/10.1080/15213269.2013.826588
Banks, A. J. (2014). The Public’s Anger: White Racial Attitudes and Opinions Toward Health Care Reform. Political Behavior, 36(3), 493–514. https://doi.org/10.1007/s11109-013-9251-3
Bartsch, A., Vorderer, P., Mangold, R., & Viehoff, R. (2008). Appraisal of Emotions in Media Use: Toward a Process Model of Meta-Emotion and Emotion Regulation. Media Psychology, 11(1), 7–27. https://doi.org/10.1080/15213260701813447
Basinger, S. (2008). The Affect Effect: Dynamics of Emotion in Political Thinking and Behavior – Edited by W. Russell Neuman, George E. Marcus, Ann N. Crigler and Michael MacKuen. Political Psychology, 29(5), 789–791. https://doi.org/10.1111/j.1467-9221.2008.00664.x
Boomgaarden, H. G., & Vliegenthart, R. (2009). How news content influences anti-immigration attitudes: Germany. European Journal of Political Research, 48(4), 516–542.
Brader, T. (2005). Striking a Responsive Chord: How Political Ads Motivate and Persuade Voters by Appealing to Emotions. American Journal of Political Science, 49(2), 388–
34 405. https://doi.org/10.1111/j.0092-5853.2005.00130.x
Brader, T. (2006). Campaigning for hearts and minds: How emotional appeals in political ads work - by Ted Brader. Journal of Communication, 56(3), 635–636.
https://doi.org/10.1111/j.1460-2466.2006.00310.x
Brader, T., Marcus, G. E., & Miller, K. L. (2011). Emotion and public opinion. In G. C. Edwards, L. R. Jacobs, & R. L. Shapiro (Eds.), Oxford Handbook of American Public Opinion and the Media. Oxford New York: Oxford University Press.
https://doi.org/10.1093/oxfordhb/9780199545636.003.0024
Brader, T., Valentino, N. A., & Suhay, E. (2008). What Triggers Public Opposition to Immigration? Anxiety, Group Cues, and Immigration Threat. American Journal of Political Science, 52(4), 959–978. https://doi.org/10.1111/j.1540-5907.2008.00353.x Bryman, A. (2016). Social research methods (Fifth edition). Oxford New York: Oxford
University Press.
Chong, D., & Druckman, J. N. (2007). Framing Public Opinion in Competitive Democracies. American Political Science Review, 101(4), 637–655.
https://doi.org/10.1017/S0003055407070554
Crigler, A. N., & Just, M. R. (2012). Measuring affect, emotion and mood in political communication. In H. Semetko & M. Scammell (Eds.), Handbook of political communication (pp. 211–225). Thousand Oaks, CA: Sage.
Decety, J. (2011). Dissecting the neural mechanisms mediating empathy. Emotion Review, 3, 92–108.
Druckman, J. N., & McDermott, R. (2008). Emotion and the Framing of Risky Choice. Political Behavior, 30(3), 297–321. https://doi.org/10.1007/s11109-008-9056-y
Druckman, J. W., & Nelson, K. R. (2003). Framing and deliberation: How citizens’
35 Eagly, A., & Chaiken, S. (1993). The Psychology of Attitudes. Fort Worth, TX: Harcourt Brase
Jovanovich.
Ellsworth, P. C., & Smith, C. A. (1988a). From appraisal to emotion: Differences among unpleasant feelings. Motivation and Emotion, 12(3), 271–302.
https://doi.org/10.1007/BF00993115
Ellsworth, P. C., & Smith, C. A. (1988b). Shades of Joy: Patterns of Appraisal Differentiating Pleasant Emotions. Cognition & Emotion, 2(4), 301–331.
https://doi.org/10.1080/02699938808412702
Forgas, J. P. (1995). Mood and Judgment: The Affect Infusion Model (AIM). Psychological Bulletin, 117(1), 39–66.
Fredrickson, B. L. (2001). The role of positive emotions in positive psychology: The broaden- and-build theory of positive emotions. American Psychologist, 56(3), 218–226.
Frijda, N. H. (1986). The emotions. Cambridge, UK: Cambridge University Press. Frijda, N. H. (1988). The laws of emotion. American Psychologist, 43, 349–358.
Frijda, N. H., Kuipers, P., & ter Schure, E. (1989). Relations Among Emotion, Appraisal, and Emotional Action Readiness. Journal of Personality and Social Psychology, 57(2), 212– 228.
Gadarian, S. K., & Albertson, B. (2014). Anxiety, Immigration, and the Search for Information: Anxiety, Immigration, and the Search for Information. Political Psychology, 35(2), 133– 164. https://doi.org/10.1111/pops.12034
Goodall, C. E., Slater, M. D., & Myers, T. A. (2013). Fear and Anger Responses to Local News Coverage of Alcohol-Related Crimes, Accidents, and Injuries: Explaining News Effects on Policy Support Using a Representative Sample of Messages and People. Journal of Communication, 63(2), 373–392. https://doi.org/10.1111/jcom.12020
Griskevicius, V., Shiota, M. N., & Neufeld, S. L. (2010). Influence of different positive
36 190–206. https://doi.org/10.1037/a0018421
Groenendyk, E. (2011). Current Emotion Research in Political Science: How Emotions Help Democracy Overcome its Collective Action Problem. Emotion Review, 3(4), 455–463. https://doi.org/10.1177/1754073911410746
Gross, K. (2008). Framing Persuasive Appeals: Episodic and Thematic Framing, Emotional Response, and Policy Opinion. Political Psychology, 29(2), 169–192.
https://doi.org/10.1111/j.1467-9221.2008.00622.x
Gross, K., & D’Ambrosio, L. (2004). Framing Emotional Response. Political Psychology, 25(1), 1–29. https://doi.org/10.1111/j.1467-9221.2004.00354.x
Harmon-Jones, C., Bastian, B., & Harmon-Jones, E. (2016). The Discrete Emotions
Questionnaire: A New Tool for Measuring State Self-Reported Emotions. PLOS ONE, 11(8), e0159915. https://doi.org/10.1371/journal.pone.0159915
Hayes, A. F. (2012). PROCESS: A versatile computational tool for observed variable mediation, moderation, and condition process modeling [White paper]. Retrieved from
http://www.afhayes.com/ public/process2012.pdf
Hayes, A. F., & Preacher, K. J. (2014). Statistical mediation analysis with a multicategorical independent variable. British Journal of Mathematical and Statistical Psychology, 67(3), 451–470. https://doi.org/10.1111/bmsp.12028
Hibbing, J. R., & Theiss-Morse, E. (1998). The Media’s Role in Public Negativity Toward Congress: Distinguishing Emotional Reactions and Cognitive Evaluations. American Journal of Political Science, 42(2), 475. https://doi.org/10.2307/2991767
Hoffman, M. L. (2000). Empathy and moral development: Implications for caring and justice. New York: Cambridge University Press.
Huddy, L., Feldman, S., & Cassese, E. (2007). On the distinct political effects of anxiety and anger. In W. R. Neuman, G. E. Marcus, A. N. Crigler, & M. MacKuen (Eds.), The affect
