The Emotional Union : Emotional Union campaigns and their effect on European’s willingness to display member-state solidarity

(1)

Running head: THE EMOTIONAL UNION 1

The Emotional Union

Emotional European Union campaigns and their effect on European’s willingness to display member-state solidarity

Sofia Vrisiis Synodinou Kiose 11571772

Master’s Thesis

Graduate School of Communication Master’s Program Communication Science

Supervisor: Dhr. Dr. Alessandro Nai Date: 29.06.2018

(2)

THE EMOTIONAL UNION 2

Abstract

The European financial crisis, followed by the immigration crisis, have led to the the rise of Euroscepticism, nationalism and a divide between European Union member-states. These developments have also led to citizens questioning and rejecting the display of solidarity to other member-states, which has a detrimental effect for the integration process. This research aimed at examining whether emotional campaigns- used by the European Union as a crisis management tool- can enhance Europeans’ willingness to display financial solidarity and solidarity for immigration, external threats and natural or man-made disasters. Moreover, the effect of Euroscepticism and identity on solidarity display and whether emotional appeals influence this relationship was also examined. These questions were investigated via an online experiment where 207 participants were exposed to four European Union video campaigns; a neutral, a hope, a pride and a fear campaign. Results indicated that a pride campaign significantly increases willingness to display all four types of member-state solidarity, a fear campaign increases willingness to display immigration member-state solidarity, while exposure to a hope campaign increases willingness to display threat member-state solidarity. Moreover, even if Euroscepticism was a strong predictor of solidarity display and identity was not, exposure to a hope appeal increases the effect of a European identity leading to positive disaster and financial solidarity display. These results not only contribute to filling a large gap in academia, but if used as a crisis management tool they can reinstall positive solidarity attitudes and mend the pan-European divide.

(3)

Emotional European Union campaigns and their effect on European’s willingness to display member-state solidarity

The issue of European solidarity, which refers to an institutionalized system of multilateral support where member-states share responsibilities and risks (Lengfeld, Schmidt & Häuberer, 2015), resurfaced as a hot topic from 2008 onward, amidst the European

economic crisis (Ciornei & Recchi, 2017). Even though the European Union (EU) and donor countries provided substantial bailout programs to many countries to avoid a Eurozone fallout (Kontochristou & Mascha, 2014), its Public Relations (PR) strategy regarding member-state solidarity proved unsuccessful (Gerrits, 2015). This occurred because creditors presented financial solidarity to debtor countries as a cold transaction, rather than as solidarity as per the founding ideals of the EU (Gerrits, 2015). Consequently, debtor countries and their citizens challenged the existence of institutionalized solidarity, EU attitudes exacerbated (Braun & Tausendpfund, 2014), and Euroscepticism and nationalism significantly increased (Bechtel, Hainmueller & Margalit, 2014).

Evidently, these developments are not only a result of political mismanagement due to lack of efficient handling (Featherstone, 2011; Hadjimichalis, 2011) but also of a poor PR strategy (Gerrits, 2015). As implementing new solidarity policies is complicated, the EU should instead undertake a crisis management approach to improve European citizens’ attitudes toward institutionalized solidarity. Citizens’ attitudes toward solidarity are essential (Lengfeld et al., 2015), as even if policy implementation does not directly depend on public support (Lengfeld et al., 2015), lack of support can lead to policy delegitimization (Van Oorschot, 2006). Therefore solidarity policies need to be accompanied by individuals’ willingness to display solidarity (Kuhn & Stoeckel, 2014; Van Oorschot, 2006).

High support of solidarity is imperative due to its importance in the integration process (Baute, Meuleman, Abts & Swyngedouw, 2017), which is at the core of the EU as an

(4)

institution (Gerrits, 2015). Lack of solidarity translates into unsuccessful integration and thus unsuccessful EU processes. As citizens are generally disengaged (Strohmeier et al., 2017) and disconnected with the EU (Hooghe & Marks, 2005), a PR tool has to be applied to augment citizens’ interest and participation in order to then increase their willingness to display solidarity. Thus, a PR crisis communication campaign needs to be implemented.

Political campaigns are rife with a variety of techniques to influence constituent’s opinions on political parties and/or candidates. One of the campaigns most frequently used is emotional campaigns, which refers to the use of emotional appeals to evoke an emotional response from individuals (Brader, 2005). This occurs as emotions strongly influence cognitive considerations and rational behavior according to individuals’ respective political environment, with different emotions having distinct effects (Brader, 2005). The

emotionalization of politics is prevalent and central in political advertising (Richards, 2004) as eliciting emotions offers paramount advantages. These advantages are not only confined in behavior and opinion change (Brader, 2006; Brader, Valentino & Suhay 2008; Marcus, 2003), but also lead to an extreme increase in political interest (Brader, 2005; Chadwick, 2015; Marcus & MacKuen, 1993) and participation (Panagopoulos, 2010; Marcus, 2000), both of which suffer a large deficit within the EU. The examination of emotions in relation to EU-issues has slowly emerged in academia, treating emotions as inherent personality traits and examining how already-existing feelings toward the EU affect attitudes for referenda (Garry, 2014) or integration (Vasilopoulou & Wagner, 2017). What has not been examined before however, is how eliciting emotions through campaigns, regardless of what citizens already feel, can influence EU-related attitudes.

The goal of this study is to explore this unexamined topic and assess whether

communication techniques can mend this Pan-European problem and improve the willingness of Europeans to display member-state solidarity holistically, namely solidarity for finance,

(5)

immigration, threats and disasters. It should be noted that the interest does not lie on whether solidarity within the EU exists, but rather how willing individuals are to support member-state solidarity. The imperative focus is solidarity attitudes, not the concept of solidarity itself. Thus, the research question is:

“What is the impact of different types of emotional campaigns on individuals’ willingness to display member-state solidarity?”

Overall, It is hypothesized that emotional campaigns will lead to higher willingness to display solidarity compared to a dispassionate campaign. Moreover, since emotions’ impact highly depends on predispositions, it is also hypothesized that emotional appeals will moderate the effect of Euroscepticism and identity on solidarity display.

Using evidence from an online experiment (N = 207), this research fills a surprisingly large gap in academia. The first lies in examining the effect of emotional appeals on

willingness to display solidarity, and specifically for crisis management purposes. Even though emotional appeals are vastly used in corporate (Kim & Cameron, 2011) and

governmental crisis management campaigns (Schneider & Hwang, 2014) and the influence of emotions on electoral attitudes is undeniable (Brader, 2006), no research on the influence of emotional appeals on solidarity attitudes as a crisis management tool has been conducted hitherto. Additionally, another gap that this paper fills is the empirical examination of solidarity as a multidimensional concept, instead of solely examining financial solidarity or bailout support (Bechtel et al., 2014; Lengfeld et al., 2015). In this study, the examination lies on citizens’ willingness to display member-state solidarity for finance, immigration, external threats and disasters.

However, the importance of this research is not only confined to the academic sphere. This examination also has paramount practical implications, as improving solidarity attitudes can enhance the EU’s accountability (De Vreese & Boomgaarden, 2005), and its institutional

(6)

legitimacy (Kuhn & Stoeckel, 2014) thus facilitating integration (Bechtel et al., 2014). Finally, this study can serve as a crisis management tool, where the EU can elicit public policy support and maintain strong intermember relationships, by knowing what emotional campaigns to launch in relation to their target audience and their demographic and

psychographic characteristics.

In the following sections the theoretical framework concerning solidarity and emotions will be provided along with the hypotheses. Next, the sample, methodology and operationalization will be described. This study will conclude by providing the results and discussion, along with limitations and future research recommendations.

European Member-state solidarity

Solidarity is an institutionalized system providing multilateral support, which requires a network of mutual understanding, responsibility, reciprocity and dependency among all EU member-states (Ciornei & Recchi, 2017). Solidarity is one of the cornerstones of the EU, as it is the foundation of the Integration process (Ciornei & Recchi, 2017), which is defined as “a process of boundary redrawing, transferring social and welfare competences to the European level” (Baute et al., 2017 p.4). This is evident by the frequency by which it is mentioned in EU documents and treaties, such as the Single European Act (1986), the Maastricht Treaty (1992) and the Treaty of Lisbon (2006) (Sangiovanni, 2013).

The focus of this paper does not lie on whether European member-state solidarity exists, but rather on the willingness of Europeans to display solidarity. As mentioned before, willingness to display solidarity is essential (Lengfeld et al., 2015) as public opposition might lead to policy delegitimization (Van Oorschot, 2006). Thus, solidarity support must occur on the national and supranational levels, as well as on the individual level (Lengfeld et al., 2015).

(7)

In sum, member-state solidarity is a governing principle for intermember relationships (Sangiovanni, 2013) for four dimensions; assistance for disasters, external threats, financial (Gerrits, 2015) and immigration solidarity (Luedtke, 2005). The first

dimension refers to financial solidarity, which entails support of member-state sovereign debt (Ciornei & Recchi, 2017) through redistribution of financial capital for economic cohesion (Baute et al., 2017; Sangiovanni, 2013). Financial solidarity amidst the European economic crisis, was the trigger for financial bailout opposition (Bechtel et al., 2014) which led to a North-South divide (Giannakopoulos, 2017).

The second dimension refers to immigration solidarity, which entails assisting member-states with intake and redistribution of immigrants or refugees (Giannakopoulos, 2017; Luedtke, 2005). While immigration solidarity is vital for maintaining free internal and protected external borders (Luedtke, 2005), many Eastern European countries have restricted (Luedtke, 2005) or have not met their obligations for refugee and immigration intake despite the Dublic Protocol (Giannakopoulos, 2017). This lack of solidarity display has created a supranational problem and an East-West divide (Sert & Türkmen, 2017).

The third dimension originated from the Article 42.7 of the European Union Treaty called the “Mutual Defense Clause” calls for external threat solidarity (Gerrits, 2015). This solidarity dimension has resurfaced in the foreground due to the potential geopolitical crises Europe could face and calls for assisting another member-state when an external force threatens its territorial sovereignty (Boin, Ekengren & Rhinard, 2013).

The last dimension originated from the Article 222 of the Lisbon Treaty calls for solidarity toward another-member state facing a man-made or natural disaster (Gerrits, 2015). This solidarity type is very important as cooperation and assistance between counties is facilitated and respective national interests are aligned (Boin et al., 2013).

(8)

Emotional appeals

While all solidarity dimensions are vital for member-state cooperation and for the integration process, no empirical examination of Europeans’ willingness to display these solidarity forms exists, except from the financial dimension. This is very surprising as increasing willingness to display member-state solidarity is imperative amidst EU divide (Giannakopoulos, 2017), disengagement (Strohmeier et al., 2017) and disconnection from the EU (Hooghe & Marks, 2005). The power of emotional appeals in changing individuals’ attitudes has been acknowledged, evidently by their frequent use in political communication (Brader, 2006). However, surprisingly insufficient research has been conducted on the impact of emotions on EU-sphere attitudes, except from the studies of Garry (2014) and

Vasilopoulou and Wagner (2017), both of which treat emotions as inherent personality traits, instead of as a reaction to a campaign. Inspired by the effects of emotions in the electoral realm, it is expected that emotional campaigns will not only be effective in shaping voting-related attitudes (Brader, 2005, 2006), but also in increasing individuals’ willingness to display solidarity.

The emotional appeals that will be examined in this research are fear, hope and pride, as there are reasons to believe that these emotions could be effectively used as a PR crisis management tool by the EU. The rationale for the selection of these specific emotions is firstly, that both negative and positive valence emotions should be examined in relation to the unexamined topic of European solidarity. Moreover, since the vast literature on emotions revolves around the Affective Intelligence Theory (AIT) and the effects of enthusiasm, anxiety and anger, those emotions or their subdimensions should be used. Anger is rarely used in crisis management, as it is a crisis consequence, not a technique for crisis resolution (Coombs & Holladay, 2011) and it was thus disregarded. Fear was selected as a negative valence emotion instead of anxiety, as presenting a relevant and important issue in a fear/loss

(9)

frame has the ability to motivate action in response to a crisis message (Coombs & Holladay, 2011). Moreover, pride and hope were selected instead of enthusiasm because of their distinct and unexamined effects. Pride, which is vastly used in campaigning (Brader, 2006), is a clear subdimension of enthusiasm, but with a unique ability to strengthen community-belonging feelings (Brader, 2006). Lastly, hope was selected as another positive emotion, because it is also vastly used in campaigning (Brader, 2006) and because of its distinct effects from enthusiasm due to its relationship with fear (Chadwick, 2015), which will be discussed below.

Emotions are perceived as a response to an external stimulus, while emotional appeals are a communication tool in order to evoke an emotional response from individuals (Brader, 2005). This discrete reaction to cues (Brader, 2006) can be simultaneously experienced as negative or positive emotions (valence) with varying intensity (arousal) (Steenbergen & Ellis, 2006) After exposure to a campaign containing emotional appeals, individuals process this information and react according to the stimulus. The way that individuals react is

examined by the Affective Intelligence Theory (AIT) which posits that two emotion systems exist; the Disposition and the Surveillance system (Marcus, Neuman & MacKuen, 2000).

The disposition system provides feedback on how well a behavior is handled; the success of a learned behavior will be perceived as enthusiasm, while failure of said behavior will be perceived as lack of enthusiasm or sadness (Marcus, 2002, 2003). When a behavior is successfully handled, enthusiasm is elicited and reliance on established predispositions is facilitated (Marcus et al., 2000). The second system, the surveillance system, checks the environment for a potential threat; if nothing threatening is detected, the system allows for the continuation of ongoing activities which will be experienced as calmness or as absence of anxiety (Marcus, 2002, 2003). However, when an unfamiliar situation is detected, the system blocks activity and transfers attention to the invasion (Marcus, 2002; Marcus & MacKuen,

(10)

1993). The feedback, which will be perceived as anxiety, enhances learning and overturns reliance on predisposition (Marcus, 2002, 2003). While fear, hope and pride do not belong in the original AIT, they are subdimensions of the original emotions and it is argued that their impact is similar to them. The specific relationships of each examined emotion with the original AIT emotions are discussed below.

Pride, a positive-valence emotion, is considered a subdimension of enthusiasm originating from the disposition system (Marcus & MacKuen, 1993; Ridout & Searles, 2011) that encourages norm-compliant behaviors and reliance on predispositions (Panagopoulos, 2014). Exposure to pride has a mobilizing effect (Flam, 2007) similar to the effect of enthusiasm (Brader, 2005). However it has a distinct difference from enthusiasm, which lies in its remarkable influence on collective identities and sense of belonging (Brader, 2006; Ridout & Searles, 2011). The effects of pride are overall quite underexamined, and while it is expected that pride will operate similar to enthusiasm, it is not known which solidarity dimensions pride will affect the most. It can, however, be argued that pride will not have an equal effect for all dimensions and that it is more likely to affect the least controversial and most consensual dimensions. This occurs as Europeans will probably be already inclined to display solidarity for the least controversial issues compared to controversial ones, in

accordance to their norms and predispositions. The issues that everyone agrees on regardless of political ideology, are called valence issues, contrary to position issues where support is contingent on partisanship or ideology (Green, 2007). Enthusiasm is connected with supporting policies without employing cognitive considerations- thus relying on

predispositions- and these policies are often valence issues (Lilleker, 2014). While financial and immigration solidarity are controversial issues (Kontochristou & Mascha, 2014; Luedtke, 2005), helping another member-state facing a sovereignty threat or a disaster can be

(11)

considered as valence issues that could be unanimously supported. Therefore, as pride is connected to enthusiasm, it is expected that:

H1: Willingness for solidarity display will be higher when participants are exposed to a pride appeal compared to the neutral campaign, especially for threat (H1a) and disaster (H1b) solidarity.

Fear belongs in the negative valence emotions and is considered a subdimension of anxiety (Marcus et al., 2000). Fear enables attitude-change as it makes individuals stop their ongoing behavior and make decisions based on new incoming information (Marcus & MacKuen, 1993). In the political electoral context, this new incoming information can be issue proximity or party evaluations, which supersede predispositions like partisanship or ideology influencing voting intentions (Marcus et al., 2000; Marcus & MacKuen, 1993). While voting intentions is not the same as solidarity attitudes, it is hypothesized that fear will operate in the same manner for solidarity display. In terms of which solidarity dimension fear will impact more, it can be argued that its impact will be higher for the more high-salience solidarity issues. Even though higher issue-salience leads to less attitude-change, in the presence of anxiety resistance is decreased and attitude-change is facilitated (Nai, Schemeil & Marie, 2017). As fear is considered an anxiety subdimension, it is expected that it will be more successful in improving willingness to display financial and immigration solidarity, as they are extremely salient issues within the EU (Kontochristou & Mascha, 2014; Luedtke, 2005) compared to threat and disaster solidarity. Therefore it is expected that:

H2: Willingness for solidarity display will be higher when participants are exposed to a fear appeal compared to the neutral campaign, especially for immigration (H2a) and financial (H2b) solidarity.

Hope belongs in the positive valence emotions (Dillard & Pfau, 2002) and is considered to be a subdimension of enthusiasm (Chadwick, 2015) that leads to increased

(12)

motivation for action (Volkman & Parrott, 2012). Even though hope strengthens feelings of well-being like enthusiasm (Druckman & McDermott, 2008), it usually emerges when a situation is perceived as frightening (Chadwick, 2015). Therefore, hope might both allow reliance on predisposition, as a subdimension of enthusiasm and also motivate alternative action, similar to how fear operates (Chadwick, 2015). As mentioned above, enthusiasm is connected to valence issue support (Lilleker, 2014) while fear, originating from anxiety, allows attitude-change for salient issues (Nai et al., 2017). Thus, as hope is connected with both enthusiasm and fear, it is expected that it will have an overcompassing effect influencing all solidarity dimensions. Its origin from enthusiasm will affect the valence dimensions, while its relationship to fear will allow attitude-change for the salient dimensions. It is thus

expected that:

H3: Willingness for solidarity display will be higher when participants are exposed to a hope appeal compared to the neutral campaign, for the threat (H3a), disaster (H3b), financial (H3c) and immigration (H3d) solidarity dimensions.

Predispositions

Even though emotional appeals can increase solidarity display, solidarity is not only dependent on emotions but also on Europeans’ predispositions, in this case their identity and level of Euroscepticism. The opinion of citizens toward the EU as a political institution (Ciornei & Recchi, 2017; Van Oorschot, 2006) and feelings of identity (Ciornei & Recchi, 2017; Gerrits, 2015) greatly influence solidarity support. Euroscepticism, which is defined as popular opposition for EU integration, is evident by negative EU attitudes (Boomgaarden, Schuck, Elenbaas & De Vreese, 2011), distrust and dissatisfaction with the EU (Van Elsas, Hakhverdian & Van der Brug, 2016).

Euroscepticism is positively associated with anti-immigration attitudes (De Vreese & Boomgaarden, 2005) and opposition to European economic governance (Kuhn & Stoeckel,

(13)

2014). Similarly, strong identification with national identities is associated with low integration support (Hooghe & Marks, 2005, 2009; Kriesi & Lachat, 2004), immigration policies opposition (Luedtke, 2005) and opposition of European economic governance (Kuhn & Stoeckel, 2014). On the other hand, identification with a European identity is strongly associated with positive member-state solidarity attitudes (Ciornei & Recchi, 2017; Gerrits, 2015) and support for bailout measures (Bechtel et al,, 2014).

While it is evident that Euroscepticism and identity have an effect on solidarity display, the impact of emotions is closely related to and can influence individuals’ predispositions. The direct effect of emotions as well as their moderating effect has been recognized in academia (Brader, 2006). Fear reduces the effect of partisanship and other predispositions on voting attitudes, while pride, as an enthusiasm subdimension, enables reliance on such predispositions (Marcus & MacKuen, 1993). The effect of hope is

unexamined and consequently less clear, but due to its connection with both AIT systems, it can be argued that it will increase reliance on predispositions for the valence issues -like enthusiasm- and decrease reliance for the high-salience issues -like fear. Inferring that the role of Euroscepticism and identity on solidarity display is similar to that of partisanship on voting, it is expected that emotional appeals will moderate the effect of Euroscepticism and identity on solidarity display.

H4a: Pride will increase the effect of Euroscepticism and identity on individuals’ willingness to display all solidarity dimensions.

H4b: Fear will decrease the effect of Euroscepticism and identity on individuals’ willingness to display all solidarity dimensions.

H4c: Hope will increase the effect of Euroscepticism and identity for the threat and disaster dimensions and decrease it for the immigration and financial dimensions.

(14)

Figure 1. Conceptual model

Methodology

The purpose of this study is to examine whether exposure to emotional appeals can increase the willingness of Europeans to display member-state solidarity, and whether emotions can influence the effect of Euroscepticism and identity. This question will be answered via an online experiment where participants are randomly exposed to different emotional campaigns and are then asked to indicate how willing they are to display solidarity for another state facing a financial, immigration, external threat or disaster crisis.

Sample

The questionnaire was created in Qualtrics and was available for completion from 12.05.2018 to 05.06.2018. During this period, a convenience and snowball sample was used. The convenience sample was recruited in two ways. The first was by contacting

acquaintances through Facebook, providing the link of the questionnaire and a brief description of the study. If these participants did not complete the questionnaire after three days, a reminder was sent. Secondly, participants were recruited through various Facebook survey pages, where students complete surveys for academic purposes. Concerning the snowball sample, the acquaintances contacted were asked to distribute the questionnaire to

(15)

other potential participants who fulfill the criteria of selection. The criteria of selection are that participants required to be at least 18 years old and EU nationals.

The final sample size consisted of 233 participants, however 26 participants were deleted, as they did not correctly answer both attention checks, concluding to N = 207. This resulted to a slightly uneven distribution of participants across conditions: neutral condition (N = 53), pride condition (N = 58), fear condition (N = 44) and hope condition (N = 52). The sample consists of more female (N=134) than male participants (N=72) with an average age of M = 28.80 (SD = 8.35; age range: 18-67 years). Concerning nationalities, participants came from 26 of the 28 EU countries, and mostly from Greece (N = 78), Netherlands (N = 29), UK (N = 16) and Germany (N = 14). In terms of education the sample is highly educated as most participants have completed a Bachelor’s (N = 115) or Master’s degree (N = 65), while only (N = 25) do not hold a university degree. The sample was somewhat politically interested (M = 3.43, SD = 1.20) and slightly right-wing (M = 4.63, SD = 2.01).

Procedure

A true between-subjects experimental design was employed, with the level of emotional campaign (fear, hope, pride, no emotion) as the between-subjects factor. As mentioned above, the questionnaire was created in Qualtrics and was distributed through an anonymous link. When participants entered the study, they first were presented with the introduction, purpose of the study and criteria of eligibility for participation. Then they were informed that participation was fully anonymous and voluntary and were asked to provide their participation consent. Demographic characteristics were collected followed by

measuring EU and identity attitudes. Then participants were randomly assigned to one of the three emotional campaigns (experimental groups) or the neutral campaign (control group) and received the stimulus. After exposure to the stimulus, the manipulation and attention checks were conducted, concluding with measuring participants’ willingness to display

(16)

solidarity. Lastly, participants were debriefed about the real purpose of the study and were informed that the video campaigns were created for the study’s purposes (see Appendix A). Stimulus

The stimulus consists of four videos, three experimental conditions; fear, hope and pride appeals and one neutral control condition. Videos were selected as the manipulation method firstly because EU institutions predominantly launch video campaigns and secondly because of the visual and auditory power of a video to elicit emotions (Brader, 2006). The videos were presented as a European Commission campaign launched to promote unity in Europe. The script was inspired by a letter sent by the President of the European Council, Donald Tusk, to the 27 EU member-states, excluding UK, for the 2017 Malta Summit urging countries to reunite and revive the integration ideal (European Council, 2017).

Since manipulation standardization is imperative in experiments to ensure that no other factor can influence the results except the manipulation, the four different videos were created with nearly identical text and image sequences. However, all videos had slight differences to elicit the different emotions, like different text, images, tone of voice and different music (Brader, 2006; Albertson & Gadarian, 2016). The fear campaign was presented as fear of losing unity within the EU, featuring a more ominous music and more violent scenes. The pride campaign was presented as being proud of sustaining unity within the EU despite difficulties and featured a more epic music with positive images. The hope campaign was presented as hoping that unity will be sustained despite difficulties, and featured the same visuals as the pride condition, but with a more uplifting music. The full scripts and links of the videos are available in Appendix B.

Pretest

A pretest was conducted with a convenience sample (N = 33) to examine whether the emotional campaigns successfully elicited the respective emotions. Participants were

(17)

presented with all four videos and asked: “To what extent did the video make you feel…?”. Six emotions were provided: anxiety, fear and anger (negative valence) and pride, hope and enthusiasm (positive valence) measured on a 5-point scale from not at all (1) to very much (5). The means of the participants were then calculated to determine the effectiveness of the manipulations. The neutral video was perceived quite differently than expected, as

participants felt the positive valence emotions quite highly. Therefore the video was altered; the music was retained- as no campaign would realistically have no music- however the volume was lowered (For pretest results, see Table 1, Appendix C).

Manipulation check

The manipulation check was presented in the form of cued recall after the completion of the posttest, following other studies examining emotional effects, such as those of Brader (2005, 2006). The scale used is adjusted from the “Modified Differential Affect Scale” (M-DAS) (Wirtz, 2018). Although only the three emotions are of interest for the manipulation check, three additional emotions were included resulting in anxiety, fear and anger (negative valence) and pride, hope and enthusiasm (positive valence). The reason behind this is to preserve the study’s realism and to avoid prompting participants.

Participants were asked the question: “To what extent did the video make you feel?” followed by the six emotions measured on a 5-point scale from not at all (1) to very much (5). Overall, it seems that compared to the pretest, the neutral video has lower means, with scores indicating that participants did not feel, or felt the indicated emotions a little. To evaluate the effectiveness of the manipulation and whether it significantly influenced participants, nine independent sample t-tests were conducted with all experimental conditions versus the neutral conditions as the independent variable and the three emotions elicited as the dependent. Concerning the emotion of pride, it was successfully manipulated in the pride condition, t(109)= -3.88, p < .001, however the hope condition participants also felt pride

(18)

t(103) = -4.75, p < .001. For the fear emotion, although fear condition participants indicated that they only felt a little to moderate fear, the manipulation was successful t(70.94) = -4.16, p < .001, however hope participants also significantly felt fear t(85.08) = 2.22, p = .029. Finally, concerning the hope emotion, the manipulation was successful for the hope condition participants t(103)= -6.50, p < .001, however those exposed to the pride condition also significantly felt hope t(109)= -2.88, p = .005. Overall, the translation between the emotions elicited and their respective video, and thus the manipulation, was successful. However, in some cases emotions were also felt in other experimental conditions than intended, which could lead to confounding results.

Table 2.

Means and Standard Deviations of Manipulation Check

Fear Anxiety Anger Pride Hope Enthusiasm

Neutral M = 1.58 SD = 0.89 M = 1.64 SD = 0.90 M = 1.64 SD = 0.90 M = 2.09 SD = 1.18 M = 2.28 SD = 1.25 M = 2.15 SD = 1.17 Pride M = 1.41 SD = 0.75 M = 1.52 SD = 0.82 M = 1.45 SD = 0.94 M = 3.00 SD = 1.27 M = 2.98 SD = 1.31 M = 2.88 SD = 1.23 Fear M = 2.55 SD = 1.37 M = 2.52 SD = 1.19 M = 2.14 SD = 1.11 M = 2.27 SD = 1.15 M = 2.41 SD = 1.23 M = 2.20 SD = 1.17 Hope M = 1.27 SD = 0.53 M = 1.38 SD = 0.53 M = 1.31 SD = 0.70 M = 3.19 SD = 1.19 M = 3.81 SD = 1.16 M = 3.23 SD = 1.11 Note. Experimental Treatments are in the rows and measurement of emotions in columns. N = 33

(19)

Measurements

Dependent variable

The dependent variable; willingness to display member-state solidarity, was

operationalized in four dimensions; financial, immigration, threat and disaster solidarity. The measurement scale for financial solidarity is borrowed by Baute et al., (2017) and is

comprised of three items measured on a 5-point scale ranging from strongly disagree (1) to strongly agree (5). Two items examine financial solidarity attitudes and one item examines attitudes on tax money redistribution, which was reverse-coded so that agreement indicated high willingness for solidarity display. Another item was included to further tap into financial assistance attitudes borrowed from Lengfeld et al. (2015) with the same measurement scale.

A principal component analysis (PCA) was conducted, indicating that all four items loaded on one factor, with an eigenvalue of 1.96 explaining 49.08% of the variance with a moderate reliability of Cronbach's alpha = .64. The new variable created was labelled “Financial solidarity” (M = 3.70, SD = 0.69).

Moreover, since no measurement scales exist for the three other dimensions, three respective scales were created. The scales were adapted from the Lengfeld et al., (2015) item, examining the desirability of the participant’s country to assist another member-state facing a crisis of immigration influx, external threat or natural or man-made disaster. Each scale consisted of two items, except from the immigration scale, which consisted of three items. This additional item tapped into immigrant relocation, as this was a major issue for some EU countries (Giannakopoulos, 2017). All scales tapped into direct (financial) and indirect (human or other resources) assistance, and were measured on a 5-point scale ranging from strongly disagree (1) to strongly agree (5). To determine if all items measure the same concepts, three principal component analyses (PCA) were conducted.

(20)

The external threat items loaded on one factor with an eigenvalue of 1.70 explaining 84.92% of the variance with a high reliability of Cronbach's alpha = .82. The new variable created was labelled “External threat solidarity” (M = 3.63, SD = 1.02).

The disaster items also loaded on one factor with an eigenvalue of 1.66 explaining 83.13% of the variance a good reliability of Cronbach's alpha = .78. The new variable created was labelled “Disaster solidarity” (M = 4.47, SD = 0.63).

The immigration items also loaded on one factor with an eigenvalue of 2.05

explaining 68.22% of the variance with good reliability of Cronbach's alpha = .74. The new variable created was named “Immigration solidarity” (M = 4.02, SD = 0.78).

Predispositions

The identity scale is borrowed from Boomgaarden et al (2011) and is comprised of five items measured on a 5-point scale ranging from strongly disagree (1) to strongly agree (5). These items examine among others what the EU means to the participant or if the participant is proud to be European. A principal component analysis (PCA) was conducted, revealing that the items loaded on one factor with an eigenvalue of 3.18 explaining 63.68% of the variance with a very good reliability of Cronbach’s alpha = .85. The items were then combined into a new variable, which was recoded into a dummy variable using the mean (M = 3.42, SD = 0.88). Values below the mean indicate “Nationalist Identity” (N = 106) while values ranging above the mean indicate “European Identity” (N = 101).

Attitudes toward the EU is a multidimensional concept, therefore it is imperative that it is measured with many items in order to fully gauge it (Boomgaarden et al, 2011). Thus, Euroscepticism- a dimension of EU attitudes- was measured with two scales. The first is comprised of seven items measuring support for EU integration borrowed from Lubbers and Jaspers (2011) with one additional item on EU trust, measured on a 5-point scale ranging from strongly disagree (1) to strongly agree (5). The second scale is borrowed from

(21)

Boomgaarden et al., (2011) and is comprised of five items examining EU attitudes and benefits of being part of the EU. The items were measured on a 5-point scale ranging from strongly disagree (1) to strongly agree (5).

Since the Lubbers and Jaspers (2011) scale measures integration support, the items “I am a proponent of further EU integration” and “Because EU countries have much in

common, integration is good” were reverse-coded, so that a high value indicated Euroscepticism. The Boomgaarden et al., (2011) scale measuring EU attitudes was also reverse-coded, so that high values indicated Euroscepticism. A Principal component analysis (PCA) was conducted, indicating that these thirteen items loaded on three factors. The first factor consists of all Boomgaarden et al., (2011) items, with an eigenvalue of 5.17 explaining 39.86% of the variance. Six of the seven items from the Lubbers and Jaspers (2011) scale loaded on the same factor with the additional item for EU trust, with an eigenvalue of 1.46 explaining 11.34% of the variance. The remaining Lubbers and Jaspers (2011) item examining national sovereignty loaded on a third factor, with an eigenvalue of 1.31 explaining 8.70 % of the variance. However, the reliability analysis of all thirteen items is very strong, with Cronbach’s alpha = .85. All items were combined into a new variable, which was then dichotomized using the mean (M = 2.54, SD = 0.66) into a new variable. Values below the mean indicate lack of euroscepticism, labelled “Europhile” (N = 122) while values above the mean indicate a “Eurosceptic”attitude (N = 85). For all the measurement items see Appendix D.

Randomization check

In order to determine whether participants were randomly assigned to the four conditions successfully, multiple randomization checks were conducted. The randomization of gender and education was examined by conducting two chi-square tests with gender and education as the dependent variables and the experimental conditions as the independent. The

(22)

results indicated that the randomizations were successful both for gender (χ2(6) = 10.04, p = .123) and education (χ2(9) = 8.89, p = .451) as the relationships were insignificant. Moreover, the randomization of other variables was also examined, by conducting multiple one-way ANOVAs, with said variables as the dependent variable and the experimental conditions as the independent. The randomization was successful for age F(3,203) = 1.04, p = .375, political interest F(3,203) = 0.85, p = .467 and political ideology F(3,203) = 0.97, p = .406.

Results

Το test the first hypothesis, two independent sample t-tests were conducted to investigate whether willingness to display threat (H1a) and disaster (H1b) member-state solidarity is higher when exposed to a pride campaign compared to a neutral campaign.

In regard to the threat solidarity, the Levene’s test for equality of variance for was not significant, F = 1.79, p = .184. The results revealed that the participants exposed to the pride appeal were more willing to display solidarity for external threats (M = 3.94, SD = 0.93) than those exposed to the neutral campaign (M = 3.33, SD = 1.05), t(109)= -3.24, p = .002, d = 0.62. Therefore, H1a is confirmed. For the disaster solidarity, the Levene’s test was significant, F = 10.92, p <.001. Participants exposed to the pride appeal were more willing to display disaster solidarity (M = 4.59, SD = 0.51) than those exposed to the neutral campaign (M = 4.26, SD = 0.72), t(92.26) = -2.76, p =.007, d = 0.53, confiriming H1b.

The other types of solidarity were also examined in relation to the effects of a pride appeal. Concerning financial solidarity, the Levene’s test was not significant, F = 0.00, p = .993. Participants exposed to the pride appeal were significantly more willing to display financial solidarity (M = 3.86, SD = 0.69) than those exposed to the neutral campaign (M = 3.45, SD = 0.69), t(109) = -3.14, p =.002, d = 0.59. Concerning immigration solidarity, the Levene’s test was not significant, F = 0.31, p = .580. Participants exposed to the pride appeal

(23)

were significantly more willing to display immigration solidarity (M = 4.13, SD = 0.76) than those exposed to the neutral campaign (M = 3.75, SD = 0.83), t(109) = -2.53, p =.013, d = 0.48. Contrary to original expectations, exposure to a pride appeal can significantly improve willingness to display all solidarity dimensions, thus H1 is rejected.

To test the second hypothesis, two independent sample t-tests were conducted to examine whether willingness to display solidarity is higher when exposed to a fear appeal compared to neutral campaign for the immigration (H2a) and financial (H2b) dimensions.

Concerning immigration solidarity, the Levene’s test for equality of variance was not significant F = 2.13, p = .148. Participants exposed to the fear campaign were more willing to display immigration solidarity (M = 4.13, SD = 0.66) compared to the neutral condition participants (M = 3.75, SD = 0.83), t(95) = -2.46, p = .016, d = 0.51, confirming H2a. However, the fear appeal did not significantly increase financial solidarity display, p = .486, disproving H2b. The other types of solidarity were also examined in relation to the effects of a fear appeal, however, the effect of fear, both for disaster (p = .056) and threat solidarity (p = .613) was non-significant. Therefore, a fear appeal does not significantly improve willingness to support any solidarity dimension, except immigration, partially confirming H2.

Concerning H3, four independent sample t-tests were conducted to determine whether participants have a higher willingness to display solidarity about all member-state

dimensions, when exposed to the hope campaign compared to the neutral condition.

For displaying threat solidarity (H3a), the Levene’s test was not significant F = 2.35, p = .128. Participants exposed to the hope campaign were significantly more willing to display threat solidarity (M = 3.76, SD = 0.87) compared to the neutral campaign participants (M = 3.33, SD = 1.05), t(103) = -2.28, p = .025, d = 0.45. Therefore, H3a is confirmed. However, the hope appeal did not significantly affect solidarity display for the remaining

(24)

three dimensions: disaster solidarity (p = .071), financial solidarity (p = .764) and immigration solidarity (p = .056). Therefore H3 is overall rejected.

Predispositions

In order to examine the direct effect of Euroscepticism and identity on solidarity display and the moderating effect of emotional campaigns, four multiple regressions were conducted for each solidarity dimension. Thus, four models will be presented, with each solidarity dimension as the dependent variable, Euroscepticism and identity as independent variables and their interaction effect with the experimental conditions.

The regression model for immigration solidarity is significant, F(11, 195) = 4.07, p < .001, with a weak strength, as 19 percent of the variation in solidarity display can be predicted by the independent variables (R2 = 0.19). Being Europhile significantly and

positively predicts immigration solidarity display b* = 0.40, t = 2.98, p = .003, 95% CI [0.22, 1.06], while identity does not. Moreover, no moderating effect of the emotional appeals on the relationship between Euroscepticism, identity and solidarity exists. Therefore, while lack of Euroscepticism significantly increases immigration solidarity display, identity does not, and no interaction between emotional appeals and individuals’ predispositions exists.

In regard to the financial solidarity, the regression model is significant, F(11, 195) = 6.67, p < .001, with a moderate strength, as 27 percent of the variation in solidarity display can be predicted by the independent variables (R2 = 0.27). Being Europhile significantly and positively predicts financial solidarity display b* = 0.45, t = 3.51, p = .001, 95% CI [0.28, 0.98], while having a European identity does not. Concerning the moderating effect of emotional appeals on the relationship between Euroscepticism, identity and financial solidarity no interaction effect is present except from hope in relation to identity. Hope increases the effect of identity on solidarity display b* = 0.29, t = 2.05, p = .041, 95% CI

(25)

[0.02, 1.07]. Overall, exposure to a hope appeal increases the effect of European identity for financial solidarity display.

Figure 2. Interaction effect between identity and hope appeal on financial solidarity

The regression model for threat solidarity as dependent variable is significant, F(11, 195) = 3.48, p < .001, with a weak strength, as 16 percent of the variation in solidarity display can be predicted by the independent variables (R2 = 0.16). Neither lack of Euroscepticism nor identity, predict threat solidarity display, and no moderating effect is present. Therefore, identity and Euroscepticism levels do not predict threat solidarity display and no interaction effect is present between emotional appeals and individuals’ predispositions.

The regression model for disaster solidarity as dependent variable is significant, F(11, 195) = 3.57 p < .001, with a weak strength, as 17 percent of the variation in solidarity display can be predicted by the independent variables (R2 = 0.17). Lack of Euroscepticism

significantly increases disaster solidarity display b* = 0.47, t = 3.48, p = .001, 95% CI [0.26, 0.95], while identity does not predict it. The only moderating effect is that of hope, as it increases the effect of identity on disaster solidarity, b* = 0.39, t = 2.33, p = .021, 95% CI

(26)

[0.09, 1.12]. Therefore, regarding disaster solidarity display, Euroscepticism is a significant predictor and hope increases the effect of identity.

Figure 3. Interaction effect between identity and hope appeal on disaster solidarity

These results lead to a rejection of H4a and H4b, and a partial confirmation of H4c (See Appendix E for the regression models).

Discussion

Concerning the effect of fear, it was expected that due to its ability to facilitate attitude-change of high-salience issues, (Nai et al., 2017), the high-salience financial and immigration solidarity dimensions (Kontochristou & Mascha, 2014; Luedtke, 2005) would be increased more. However the effect of fear only occured in regard to immigration and not for financial solidarity. Moreover, fear did not significantly increase low-salience valence issues either, namely willingness to display solidarity for external threat and disaster.

Additionally, hope was expected to increase all solidarity dimensions, as even if it is considered a subdimension of enthusiasm it can also motivate alternative action, similar to fear (Chadwick, 2015). Therefore, hope would increase willingness to display solidarity both for high-salience partisan issues and low-salience valence issues. This was not the case, as

(27)

hope only significantly affected threat attitudes. Therefore, no overarching and encompassing effect of hope was detected, even when hope was the most successfully elicited emotion.

The pride appeal was the most successful in increasing solidarity display, disproving original expectations. Instead of hindering attitude-change of salient issues and only affecting valence issues, as an enthusiasm subdimension, pride significantly increased all solidarity dimensions. This finding is extremely interesting, especially as pride was the only emotion that significantly increased willingness to financially assist a member-state. Thus, it seems that even though pride is considered a subdimension of enthusiasm (Marcus & MacKuen, 1993; Ridout & Searles, 2011), which would not facilitate attitude-change of high-salience issues, this is not the case. This finding can be explained due to pride’s ability to influence collective identities (Brader, 2006; Ridout & Searles, 2011). Thus, its ability to unite people and strengthen community relations makes a difference in all solidarity attitudes.

In regard to the effects of Euroscepticism and identity on solidarity and the influence of emotional appeals on this relationship, it was expected that the pride appeal would increase the effect of Euroscepticism and identity on willingness to display solidarity, as it allows reliance on predispositions (Ridout & Searles, 2011). Secondly it was expected that the fear appeal would decrease the effect of Euroscepticism and identity, as it breaks reliance on predispositions (Brader, 2005; Marcus & MacKuen, 1993). Lastly it was hypothesized that hope, as a combination of the aforementioned emotions, would increase reliance on

predispositions for the valence dimensions and decrease reliance for the high-salience dimensions.

Lack of Euroscepticism strongly influenced solidarity support, as being a Europhile led to higher levels of solidarity support for all dimensions. This finding is in line with the findings of De Vreese and Boomgaarden (2005) and Kuhn and Stoeckel (2014), where lack of Euroscepticism leads to positive attitudes for immigration or financial assistance.

(28)

However, even though higher European identity overall led to higher solidarity support, this was not significant. This finding is very surprising as it is incongruent with a vast body of literature like the studies of Ciornei and Recchi (2017), Gerrits (2015), and Kuhn and Stoeckel (2014).

Concerning the influence of emotional appeals in decreasing or increasing the importance of identity and Euroscepticism on solidarity attitudes, no appeal was successful except from hope. Fear did not significantly decrease the importance of these predispositions and pride did not increase it, disproving original expectations concerning their effects

(Marcus & MacKuen, 1993). Hope was the only influential appeal, as it led to an increase of the effect of identity for the financial and disaster dimensions. This finding also contradicts original expectations, as even though the effect of identity on the disaster dimension was expected, this was not the case for the financial dimension, where its effect should have been decreased. Thus, it seems that hope functions solely as a subdimension of enthusiasm in terms of predispositions, and that when a hope appeal exists in conjunction with a European identity both valence and high-salience issues can be affected. This is a very interesting result as if individuals that highly identify as Europeans are presented with a hope campaign, this can lead to higher willingness to assist another state facing a disaster and financial crisis. Conclusion

This research aimed at examining the effects of three emotional campaigns, namely pride, hope and fear campaigns on Europeans’ willingness to display financial, immigration, threat and disaster solidarity to assist another EU state. Moreover, Euroscepticism and identity were also expected to predict solidarity display and that this relationship would be influenced by emotional appeals.

The results, obtained via conducting an online experiment (N = 207), indicated that exposure to a pride campaign increases willingness to display all four types of member-state

(29)

solidarity. Moreover, exposure to a fear campaign increases willingness to display

immigration solidarity, while exposure to a hope campaign increases willingness to display threat solidarity. Additionally, even though Euroscepticism was a strong predictor of

solidarity display, identity was not. Finally, hope can increase the effect of a strong European identity leading to higher willingness to display disaster and financial solidarity.

This has big practical implications, as if the EU wants to launch campaigns addressing the aforementioned solidarity issues, specific emotional appeals should be used for specific issues. To increase financial and disaster solidarity, the only effective appeal is pride, while for immigration solidarity both pride and fear appeals are equally influential. Finally, for threat solidarity, both hope and pride appeals can be used, although pride is more effective. Additionally, if the EU were to launch campaigns for financial and disaster solidarity to a European identification target audience, this campaign should include a hope appeal. Finally, tailoring the campaign’s content according to the target audience’s Euroscepticism levels does not make a difference on the campaign’s impact.

This study also has significant academic implications. The first is the examination of willingness to display memberstate solidarity as a multidimensional concept. The second -and most important- is creating a bridge between two literatures; emotional campaigning -and solidarity attitudes, by posing questions unexamined hitherto. This study proved that

emotional campaigns can not only influence voting behavior (Brader, 2005), but also willingness to display solidarity.

Concerning the limitations of this study, the standardization of stimuli required in an experimental design made eliciting different emotions using almost identical scripts and visuals very challenging. This limitation might be the reason why the emotions elicited were successfully manipulated not only in the respective experimental condition, but in other conditions as well. Consequently, it is possible that results might have led to confounding

(30)

outcomes. Moreover, due to the exclusion of twenty-six respondents, the fear condition was comprised of fewer participants compared to the other conditions, which could also have affected the results. Lastly, emotions were measured via self-report measures, which can be problematic as they can lead to be biased results (Arceneaux, 2012).

For future research it is recommended that different stimuli tailored to eliciting each respective emotion could be used, instead of using manipulation standardization of

experimental designs. Moreover, a larger sample size is recommended, as many hypotheses were borderline rejected, which could not be the case with a larger sample. Additionally since attitudes toward solidarity display are very ingrained, a single exposure to emotional

campaigns might not be adequate in improving solidarity attitudes. Therefore, a longitudinal study could be conducted to examine the impact of emotional campaigns over time.

In conclusion, this study has proved that emotional appeals can increase the

willingness of Europeans to help another member-state. This finding does not only fill a large gap in academia. If this finding is used as a crisis management tool, it can be a stepping-stone in reducing hostility and divide among nations, reviving the integration process and

reinstalling the EU as a vital and important organization in Europeans’ lives.

References

Albertson, B., & Gadarian, S. K. (2016). Did that scare you? Tips on creating emotion in experimental subjects. Political Analysis, 24(4), 485-491. doi:10.1093/pan/mpw022 Arceneaux, K. (2012). Cognitive biases and the strength of political arguments. American

Journal of Political Science, 56(2), 271-285. doi: 10.1111/j.1540-5907.2011.00573.x Baute, S., Meuleman, B., Abts, K., & Swyngedouw, M. (2017). Measuring attitudes towards

Social Europe: A multidimensional approach. Social Indicators Research, 1-26. doi: 10.1007/s11205-017-1587-3

(31)

Bechtel, M. M., Hainmueller, J., & Margalit, Y. (2014). Preferences for international redistribution: The divide over the Eurozone bailouts. American Journal of Political Science, 58(4), 835-856. doi: 10.1111/ajps.12079

Boomgaarden, H. G., Schuck, A. R., Elenbaas, M., & De Vreese, C. H. (2011). Mapping EU attitudes: Conceptual and empirical dimensions of Euroscepticism and EU support. European Union Politics, 12(2), 241-266. doi: 10.1177/ 1465116510395411 Boin, A., Ekengren, M., & Rhinard, M. (2013). The European Union as crisis manager:

Patterns and prospects. Cambridge University Press.

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), 388-405. doi: 10.1111/j.0092-5853.2005.00130.x

Brader, T. (2006). Campaigning for hearts and minds: How emotional appeals in political ads work. University of Chicago Press.

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. doi: 10.1111/j.1540-5907.2008.00353.x

Braun, D., & Tausendpfund, M. (2014). The impact of the Euro crisis on citizens’ support for the European Union. Journal of European Integration, 36(3), 231-245. doi:

10.1080/07036337.2014.885751

Chadwick, A. E. (2015). Toward a theory of persuasive hope: Effects of cognitive appraisals, hope appeals, and hope in the context of climate change. Health communication, 30(6), 598-611. doi: 10.1080/10410236.2014.916777

Ciornei, I., & Recchi, E. (2017). At the Source of European Solidarity: Assessing the Effects of Cross‐border Practices and Political Attitudes. JCMS: Journal of Common Market Studies, 55(3), 468-485. doi: 10.1111/jcms.12507

(32)

Coombs, W. T., & Holladay, S. J. (Eds.). (2011). The handbook of crisis communication (Vol. 22). John Wiley & Sons.

De Vreese, C. H., & Boomgaarden, H. G. (2005). Projecting EU referendums: Fear of immigration and support for European integration. European Union Politics, 6(1), 59-82. doi: 10.1177/1465116505049608

Dillard, J. P., & Pfau, M. (2002). The persuasion handbook: Developments in theory and practice. Sage Publications.

Druckman, J. N., & McDermott, R. (2008). Emotion and the framing of risky choice. Political Behavior, 30(3), 297-321. doi: 10.1007/s11109-008-9056-y

European Council. (2017, January 31). “United we stand, divided we fall”: letter by

President Donald Tusk to the 27 EU heads of state or government on the future of the EU before the Malta summit [Press release]. Retrieved from

http://www.consilium.europa.eu/en/press/press-releases/2017/01/31/tusk-letter-future-europe/

Featherstone, K. (2011). The JCMS annual lecture: The Greek sovereign debt crisis and EMU: A failing state in a skewed regime. JCMS: Journal of Common Market Studies, 49(2), 193-217. doi: 10.1111/j.1468-5965.2010.02139.x

Flam, H. (2007). Emotions’ map: A research agenda. In Flam, H., & King, D. Emotions and social movements. Routledge.

Hadjimichalis, C. (2011). Uneven geographical development and socio-spatial justice and solidarity: European regions after the 2009 financial crisis. European Urban and Regional Studies, 18(3), 254-274. doi: 10.1177/0969776411404873

Hooghe, L., & Marks, G. (2005). Calculation, Community and Cues: Public Opinion on European Integration. European Union Politics, 6(4), 419-443. doi:

(33)

Hooghe, L. and Marks, G. (2009). A postfunctionalist theory of European integration: From permissive consensus to constraining dissensus. British Journal of Political Science, 39 (1), 1-23. doi:10.1017/S0007123408000409

Garry, J. (2014). Emotions and voting in EU referendums. European Union Politics, 15(2), 235-254. doi: 10.1177/1465116513514780

Gerrits, A. M. W. (2015). Solidarity and the European Union: From the welfare state to the Euro crisis. Shaping a different Europe: Contributions to a critical debate. Bonn: Dietz Verlag JHW Nachf.

Giannakopoulos, A. (2017). Solidarity: Sociological, Legal and Ethical Aspects of a

Fundamental EU-Principle. In Giannakopoulos, A. (Ed.), Solidarity in the European Union: Challenges and Perspectives. S. Daniel Abraham Center for International and Regional Studies, Tel Aviv University.

Green, J. (2007). When voters and parties agree: Valence issues and party competition. Political Studies, 55(3), 629-655. doi: 10.1111/j.1467-9248.2007.00671.x Kim, H. J., & Cameron, G. T. (2011). Emotions matter in crisis: The role of anger and

sadness in the publics’ response to crisis news framing and corporate crisis response. Communication Research, 38(6), 826-855. doi: 10.1177/0093650210385813

Kontochristou, M., & Mascha, E. (2014). The Euro crisis and the question of solidarity in the European Union: Disclosures and manifestations in the European press. Review of European Studies, 6(2), 50. doi:10.5539/res.v6n2p50

Kriesi, H., & Lachat, R. (2004). Globalization and the transformation of the national political space: Switzerland and France compared. doi: 10.2139/ssrn.1514444 Kuhn, T., & Stoeckel, F. (2014). When European integration becomes costly: the euro crisis

and public support for European economic governance. Journal of European Public Policy, 21(4), 624-641. doi: 10.1080/13501763.2013.867892

(34)

Lengfeld, H., Schmidt, S., & Häuberer, J. (2015). Is there a European solidarity? Attitudes towards fiscal assistance for debt-ridden European Union member states. doi: 10.2139/ssrn.2597605

Lilleker, D. G. (2014). Proximity Politics and Valence. In Political Communication and Cognition. Palgrave Macmillan, London. doi: 10.1057/9781137313430_6

Lubbers, M., & Jaspers, E. (2011). A longitudinal study of euroscepticism in the Netherlands: 2008 versus 1990. European Union Politics, 12(1), 21-40. doi:10.1177/

1465116510390062

Luedtke, A. (2005). European integration, public opinion and immigration policy: Testing the impact of national identity. European Union Politics, 6(1), 83-112. doi: 10.1177/ 1465116505049609

Marcus, G. E. (2002). The Sentimental Citizen. University Park.

Marcus, G. E. (2003). The psychology of emotion and politics. Oxford handbook of political psychology, 182-221.

Marcus, G. E., & MacKuen, M. B. (1993). Anxiety, enthusiasm, and the vote: The emotional underpinnings of learning and involvement during presidential campaigns. American Political Science Review, 87(3), 672-685. doi: 10.2307/2938743

Marcus, G. E., Neuman, W. R., & MacKuen, M. (2000). Affective intelligence and political judgment. University of Chicago Press.

Nai, A., Schemeil, Y., & Marie, J. L. (2017). Anxiety, Sophistication, and Resistance to Persuasion: Evidence from a Quasi‐Experimental Survey on Global Climate Change. Political Psychology, 38(1), 137-156. doi: 10.1111/pops.12331

Panagopoulos, C. (2014). I've got my eyes on you: Implicit social‐pressure cues and prosocial behavior. Political Psychology, 35(1), 23-33. doi: 10.1111/pops.12074

(35)

Ridout, T. N., & Searles, K. (2011). It's my campaign I'll cry if I want to: How and when campaigns use emotional appeals. Political Psychology, 32(3), 439-458. doi: 10.1111/j.1467-9221.2010.00819.x

Richards, B. (2004). The emotional deficit in political communication. Political Communication, 21(3), 339-352. doi: 10.1080/10584600490481451

Sangiovanni, A. (2013). Solidarity in the European Union. Oxford Journal of Legal Studies, 33(2), 213-241. doi: 10.1093/ojls/gqs033

Sert, D. S., & Türkmen, F. F. (2017). The EU-Turkey Refugee Deal: The “Baleful” Balance between Protection of Refugee and Human Rights and Controlled Refugee Flows. In Giannakopoulos, A. (Ed.), Solidarity in the European Union: Challenges and

Perspectives. S. Daniel Abraham Center for International and Regional Studies, Tel Aviv University.

Schneider, F., & Hwang, Y. J. (2014). The Sichuan Earthquake and the Heavenly Mandate: legitimizing Chinese rule through disaster discourse. Journal of Contemporary China, 23(88), 636-656. doi 10.1080/10670564.2013.861145

Steenbergen, M. R., & Ellis, C. (2006). Fear and loathing in American elections: Context, traits, and negative candidate affect. In Feeling politics (pp. 109-133). Palgrave Macmillan, New York. doi: 10.1057/9781403983114_7

Strohmeier, D., Barrett, M., Bora, C., Caravita, S. C., Donghi, E., Dragoti, E., ... & Rama, R. (2017). Young People’s Engagement With the European Union. Zeitschrift für Psychologie. doi: 10.1027/2151-2604/a000314

Van Elsas, E. J., Hakhverdian, A., & Van der Brug, W. (2016). United against a common foe? The nature and origins of Euroscepticism among left-wing and right-wing citizens. West European Politics, 39(6), 1181-1204. doi:

(36)

Van Oorschot, W. (2006). Making the difference in social Europe: deservingness perceptions among citizens of European welfare states. Journal of European social policy, 16(1), 23-42. doi:10.1177/0958928706059829

Vasilopoulou, S., & Wagner, M. (2017). Fear, anger and enthusiasm about the European Union: Effects of emotional reactions on public preferences towards European integration. European Union Politics, 18(3), 382-405. doi: 10.1177/

1465116517698048

Volkman, J. E., & Parrott, R. L. (2012). Expressing emotions as evidence in osteoporosis narratives: Effects on message processing and intentions. Human Communication Research, 38(4), 429-458. doi: 10.1111/j.1468-2958.2012.01433.x

Wirz, D. S. (2018). Persuasion Through Emotion? An Experimental Test of the Emotion-Eliciting Nature of Populist Communication. International Journal of

(37)

Appendix A Introduction

Dear participant,

Welcome! Thanks for your interest in this study.

Achieving European solidarity, although vital, is one of the most debatable issues due to current socioeconomic developments. In this study, we are interested to know your attitudes on European solidarity and the European Union. For you to be eligible you have to be over 18 years old and a *European Union citizen*.

You will be presented with a European Union campaign and asked to provide your attitudes along with a few personal questions and general political questions, to help us interpret the results.

It is important to note that the questionnaire is fully anonymous. We will not be able to identify the participants, so you should be free to answer according to your opinions and beliefs. Furthermore, participation is purely voluntary.

You can deny participation in the research and withdraw at any time, without having to provide a reason for doing so. After the questionnaire is completed, you have up to 7 days to withdraw the permission for the use of your data in the research.

You are welcome to contact the principal investigator in the study

(sofia.synodinoukiose@student.uva.nl) for any questions, as well as the Ethics committee for any complaint or questions about the study:

(38)

THE EMOTIONAL UNION 38 ASCoR Secretariat Ethics Committee University of Amsterdam PO Box 15793 1001 NG Amsterdam Tel: 020-525 3680 Email: ascor-secr-fmg@uva.nl

No specific knowledge on the European Union is required. The questionnaire is in English, is better displayed on computer screens and takes approximately 10 minutes to complete. Thank you in advance!

Sofia Vrisiis Synodinou-Kiose UvA sofia.synodinoukiose@student.uva.nl

Informed Consent

Please check this box to signal that you have read and understood the information provided to you concerning the scope of the research, anonymity, and the possibility to withdraw your answers at a later stage. After that, please click on the “next” ( -> ) button to start the questionnaire. You cannot come back to previously answered questions.

(39)

Debrief

This study aimed at examining how different emotional campaigns and the emotions elicited in them can affect the attitudes toward European solidarity, in relation to existing EU

attitudes.

Before concluding, we would like to emphasize that the campaign video of the European Union you were exposed to is fictional, does not necessarily represent the views of the European Union and was created for the needs of this study. Even though the information provided in the video is accurate, the video is fictional.

If you have any questions or comments for us about this experience, please leave them in the box below. Alternatively, feel free to send an email to the principal investigator of the study, at sofia.synodinoukiose@student.uva.nl