The European Union: A deliberative democracy?
The effects of deliberation on European Union attitudesby Tawfik, Tom Student Number: 6160050
Master Thesis
Political Communication
Amsterdam Graduate School of Communication Research Department of Communication Science
2
ABSTRACT
Deliberative democracy has been subject to extensive research in the last couple of decades. Idealists believe that institutionalizing deliberation can significantly improve our democracy. Due to the rise of euroscepticism the European Union (EU) follows by example through experimenting with a deliberative democratic ideal. EuroPolis, a deliberative experiment in 2009, demonstrated significant changes in deliberators’ attitudes towards the EU (becoming more pro-European). Some scholars suggest that it is not deliberation itself but contextual factors that account for these effects. This study conducts a deliberative experiment (N=65) on attitudes towards the EU in a laboratory setting with pre- and posttest measurements and a control group. This is the first deliberative experiment that includes a multidimensional concept of EU attitudes: performance, identity, affection, utilitarianism and strengthening. The results of this study contradict findings of earlier studies that point towards more positive attitude after deliberating. Instead, deliberation significantly increases negative affection. Findings of this study suggest that contextual factors might have been accountable for the positive effects of deliberation on EU attitudes. More empirical research on deliberation, in varying contexts and settings, and its effects on EU attitudes is needed to either confirm or falsify the present findings. Until proven wrong, deliberative democracy might not be the solution to overcoming euroscepticism after all.
3
INTRODUCTION
The word democracy is ancient Greek for “rule of the people” and was intended to include all citizens in a governmental process to legitimize policy-making. Democracy as we now know it was first organized in ancient Athens at around 500BC. Ordinary citizens met at the Town Hall and deliberated about politics, set the agenda and even took place in governmental positions.
Democracy within the European Union has not yet convinced “the people” of their rule. Many scholars have written about a ‘democratic deficit’ (Follesdahl & Hix, 2006; Moravcsik, 2004; Majone, 1998). The influence and legislative power of the European Parliament have
increased while voter turnouts have dropped and real ‘European’ elections do not exist. As a consequence, citizens perceive the EU as being bureaucratic, too complex, unresponsive to the public and difficult to assess its’ performance (Follesdahl & Hix, 2006). Over the last couple of years, research on EU attitudes has shifted from EU support to euroscepticism being the dominant term (Boomgaarden, Schuck, Elenbaas and de Vreese, 2011). Euroscepticism might sway public opinion against the EU and further integration (de Vries & Edwards, 2009).
In an attempt to overcome euroscepticism, the EU has shown keen interest in experimenting with a deliberative democracy. According to Fishkin (2005) deliberative democracy is based on two core democratic values – deliberation and political equality. Over the last two decades ample amounts of evidence have shown that deliberation among citizens affects their attitudes (Fishkin, 1997;2014; Himmelroos & Christensen, 2014; Sanders, 2012; Gastil, Bacci & Dollinger, 2010; Barabas, 2004).
4
The measured effects of deliberation on euroscepticism are promising. In 2009 a deliberative experiment called EuroPolis was organized and held at the European Parliament. Bernhagen and Schmitt (2014) found that knowledge was gained and changes occurred in individual electoral behavior. Furthermore, Sanders’ (2012) analysis pointed to significant changes in the deliberators’ attitudes on immigration (becoming more liberal), climate change (becoming greener) and the EU itself (becoming more pro-European).
However, it is not sure whether deliberation itself causes attitudes to change. According to Sanders (2012) the observed effects of EuroPolis might not be explained due to the
deliberation itself. Rather the style, values and outlook of those organizing the deliberation seem to be the explaining factor for the favorable attitude changes. Furthermore, Muhlberger (2005) found in an experiment that it was not the deliberation that changed the attitudes of deliberators. Instead, information that was given to both deliberators and non-deliberators seemed to be accountable for attitude changes.
Moreover, not all deliberative experiments demonstrated the favorable effects of deliberation. Mendelberg and Oleske (2000) studied two rhetorics of town hall meetings where people of different races deliberated about school desegregation. The study showed evidence for a lack of perceived equal respect and the conflict between groups maintained. Wojcieszak (2011) studied quasi-experimental data from participants in face-to-face deliberations about sexual minority rights in Poland. Findings suggest that deliberators whom held extreme views and perceived disagreement polarized on the discussed policies and on more general issues related to sexual minorities.
The present study aims to investigate whether the future of the EU lies in deliberative democracy. A democratic deficit and increasing euroscepticism have led the EU to experiment with a deliberative democratic ideal. EuroPolis has demonstrated promising
5
effects on EU attitudes however the proportion of attitude change that the deliberation itself can be held accountable for remains unclear. Moreover, several other deliberative experiments have failed to produce the acclaimed effects on attitudes. Therefore, more empirical evidence needs to be collected in order to make a case for institutionalizing deliberative democracy in Europe and thereby overcoming euroscepticism. Therefore the research question of this study is:
What are the effects of deliberation on EU attitudes?
This study will be the first to include a multidimensional concept of EU attitude, defined by Boomgaarden et al. (2011), in a deliberative experiment.
DELIBERATION
Deliberation is one of the core principles of deliberative democratic theory. The work on democratic deliberative theory has grown ever since the “deliberative turn” and it is now common place to talk and exchange ideas about deliberative democratic theory among theorists and empiricists (Thompson; 2008; Mutz, 2008; Dryzek, 2000). Deliberative democratic theory gained much attention in public opinion research because of the insight it provides on a hypothetical public, one more engaged and informed about politics than citizens in ‘reality’ are (Luskin, Fishkin & Jowell, 2002). The aim of these studies is to determine whether and how individual attitudes would change if the information and interest gap that characterizes the public discussion would be filled in.
Furthermore, deliberative democratic theory is a democratic ideal that is claimed to be more just and democratic in its way of dealing with pluralism than other models of democracy. It is anchored in conceptions of accountability and discussion and replaces a voting-centric
6
democratic theory for a talk-centric one (Chambers, 2003). A voting-centric democracy is based on interests and preferences that compete through mechanisms of aggregation. Talk-centric deliberative democracy focuses on the communicative and cognitive processes of opinion and will formation that precedes voting. According to Fishkin (2005), at the heart of deliberative democratic theory lies political equality and deliberation.
In its most rough definition, deliberation can be seen as discussion that occurs in a wide range of practices, from political conversation, to loosely structured group discussions, to formal deliberative events (Zhang, 2012). Deliberation is a communicative and cognitive process during which citizens with different views engage in political talk while keeping an open mind. According to Barabas (2004) deliberation differs from ordinary discussion in that strongly held views are being softened and they encounter different perspectives through which they learn. Based on keeping an open mind deliberation changes attitudes (Barabas, 2004).
H1: Deliberation has an effect on EU attitudes.
Gastil, Black and Moscovitz (2008) have operationalized deliberation as containing three components. A deliberative process should hold a basic decision-making function, for example by weighing pros and cons. The second prerequisite for deliberation are the democratic elements such as considering one another’s arguments, demonstrating mutual respect and creating opportunities to express different points of view. The third and final component of deliberation should be the opportunity for dialogue when different ways of speaking or reasoning occur.
7
EU ATTITUDES
Boomgaarden et al (2011) point out to the need of reviewing the conceptual approaches to EU attitudes. It does not suffice to approach the concept of an EU attitude as a one-dimensional concept. Research has shown that attitudes of citizens towards the EU are highly ambivalent (de Vries, 2013; Stoeckel, 2013) and multidimensional (Boomgaarden et al 2011). In other words, citizens may be in favor of EU policies when it comes to environmental issues or regional stability but could be opposed due to a lack of transparency or increase of migration.
Based on public opinion survey data, Boomgaarden et al (2011) find evidence for five distinct dimensions that prove the multidimensional nature of attitudes towards the EU. The first dimension is the affective dimension and concerns the emotional responses to the EU. These emotional responses refer to fear of and threat by the EU. According to Boomgaarden et al (2011) it is the first time that an emotional affective dimension has been identified in the literature concerned with EU attitudes. The second dimension holds the sense of European identity. This dimension relates to feelings of EU citizenship and the perception of a common heritage, culture and traditions shared by Europeans. The third dimension refers to the
perceived performance and the democratic as well as financial functioning of the EU and its institutions. The performance dimension is linked with perceptions of a well-functioning and transparent EU and a trustworthy European Parliament. The fourth dimension concerns utilitarian attitudes concerned with cost-benefit evaluations and general support. An example of a statement linked to the utilitarianism dimension is the attitude towards membership of the EU. The fifth and final dimension refers to the attitude towards further strengthening and integration of the EU in the future. The attitude towards European integration and its’ speed are included in this dimension (Boomgaarden et al (2011).
8
H2: The effect of deliberation on EU attitudes varies across five dimensions: affective, identity, performance, utilitarianism and strength
METHOD
Sample
Inclusion in the target population was determined by whether or not participants were eligible to vote. This meant that everyone within reach from the age of 18 and above were part of the target population. For this study a convenience sample was used due to limited time and resources. This means that participants were recruited via social networks. Data relevant to this research was obtained through pretest and posttest surveys. The minimum sample size per condition was 30 people. The target sample size would be to create a sample size as large as possible. The convenience sampling held that everything has been done to reach as much potential participants as possible. Recruitment was done through telephone calls and messages as well as word-of-mouth and communication through social media. Apart from the minimum age of 18 there was no selection process after participants agreed to participate. A total of number of 542 people were contacted, 16,97% participated (N=92) and 83,03% (N=450) people either refused to participate or did not respond. Dropout after the pretest was 29,35% because of either unfinished questionnaires or non-response at the posttest. This means that from the initial 92 participants 65 were included in this study. As the target population holds as many people possible and as diverse possible to represent a reflection of society, the completed sample deviates in countless ways. For example, 63% (N=41) of the sample was a
9
student and 53,85% (N=35) was between the age of 18-24, 46,15% (N=30) completed at least a bachelor degree in science education. This is not close to a representative sample of the target population as for example roughly 1.45% people are enrolled in science education in the Netherlands.
Procedure
After agreeing to participate, subjects were asked to fill in a survey (pretest) one day before the deliberation took place. On deliberation day 32 people were randomly assigned to one out of six groups. Apart from the pretest questionnaire there was no information about the
experiment, let alone the topic of discussion. Respondents were told that there would be a group discussion that would last between 30-60 minutes. All group deliberations have taken place in the same room with one trained, neutral moderator. Group sizes varied between four and eight participants. All participants filled in the pretest and posttest surveys in their own private time behind a computer or mobile device.
Manipulation/Independent variable (Deliberation)
The group discussions started out with a brief explanation of the purpose of the deliberation. Participants were told that the group discussion would be concerned about the EU and that the purpose of the discussion was to deliberate together on the topic. The trained moderator set out rules for discussion and made clear that there were no right or wrong answers or
viewpoints. It was the task of the moderator to make sure every participant had equal opportunity to express their views without other participants interrupting. After a round of introductions the deliberation moved to a brainstorm session, asking the participants what was in their top-of-mind-awareness when thinking of the EU. Following the brainstorm session, two topics were discussed: pros and cons of EU membership for the Netherlands and pros and cons of immigration as a result of open borders within the EU. The moderator made clear that
10
the purpose was to first elaborate on the pros before moving on to the cons. In this way participants, even when feeling skeptic about the membership of the EU, were forced to think and elaborate on possible pros of EU membership. During these deliberative sessions the moderator actively structured the deliberation process. For example, if a participant would interrupt someone else or deviate from the topic, the moderator made clear that participants should let each other finish their sentences or to save their arguments for when we arrive at the relevant topic. Furthermore, the moderator actively summarized all pros and cons to make sure all viewpoints were expressed equally. After finishing deliberation on both topics, participants were thanked for their contribution and asked to fill in the digital posttest after leaving the room. In this way, the only difference between the control group and the experimental group was the deliberation itself.
Dependent variable (EU attitudes)
The questionnaire included all five dimensions that were found in the research done by Boomgaarden et al (2011) (see table A5 in the appendix). Every dimension contained five statements of which respondents had to rate their attitude on a seven-point likert scale. Gender has been made a dichotomous variable (female=0, male=1). Due to categorized answer
possibilities, age has been made a dichotomous variable (0=older than 24; 1=18-24). Because of the convenience sample there was a dominance of students. Therefore education has been recoded into different variables in a way that ‘0’ stands for every type of diploma or degree except a scientific educational degree. Control variables such as political interest, political ideology and political knowledge have been added to the equation. Participants were asked to rate their interest in politics on a seven-point likert scale varying from very much to not at all. Political ideology, in the footsteps of Boomgaarden et al. (2011) has been rated on an 11-point likert scale with 0 being extreme left and 11 being extreme right. In line with Boomgaarden et al (2011), political ideology was recoded into two dummy variables being left or right. For
11
both dummy variables the 4 most extreme scale points were coded ‘1’ while all other options were coded ‘0’. Finally, political knowledge has been measured by asking seven questions about national and European politics. A timer to the survey was added to prevent participants searching for additional information on the internet, phone or via friends to answer the questions. Each participant had 20 seconds per question. When the time was up, the
questionnaire automatically continued to the next question. Every good answer was coded as a 1 and don’t knows or wrong answers were coded 0. The total sum of good answers formed a scale for political knowledge.
RESULTS
When the same group of people are pretested and post-tested on one or more dependent variables, a General Linear Model Repeated Measures is recommended. Especially when interested in the performance over time on these dependent variable. In this study, two groups of people are being measured twice over time. One group is not being exposed to the
manipulation, deliberation, and the other group is. Therefore a repeated measures mixed ANOVA, or MANOVA, is the way to go. Difficulty with this analysis is that there is a long list of assumptions that the data should live up to, in order for a MANOVA to be valid. Dependent variables should be measured at the interval or ratio level (i.e. they are
continuous). The independent variable should consist of two or more categorical independent groups. An independence of observations is necessary. This means that different groups hold different participants. There should be no participant in more than one group. Furthermore, an adequate sample size is needed. For MANOVA, more cases in each group than the number of dependent variables should be in place. In this study, the pretest and posttest scores on five dimensions result in 10 different dependent variables. The sample consists of 65 participants
12
so there should be no problem here. Then, there should be no univariate or multivariate outliers. In order to check this, z-scores have been made and Mahalanobis distance has been measured. In total there were three univariate outliers that exceeded the maximum z-score allowed. After locating the outliers, decision has been made to include these in the
MANOVA. This decision has been made due to the limited amount of participants in both conditions, 32 versus 33. These outliers did not exceed the z-score by a large amount and were therefore acceptable. No multivariate outliers were found. Furthermore, linear
relationships and homogeneity of variance-covariance matrices was present. Levene’s test for homogeneity (see table A2 in the appendix) was indicated that the assumption for between-group homogeneity is satisfied across the condition for all dependent variables (significance greater than .05). Multicollinearity varied among the variables. The critical value for
multicollinearity would be 3 for VIT. Repeated linear regressions on all dependent variables showed variables being either just up or down. All nine assumptions that need to be met in order to calculate valid results are confirmed. Therefore repeated measures MANOVA analysis was conducted.
Repeated-measures MANOVA analyses showed that there was no significant multivariate effect for the groups (V = .101, F(5,59)=1.329, not significant). This means that there was no significant difference between deliberators and non-deliberators on the scores on EU attitudes. However, a significant effect for time (V=.214, F(5,59)=3.217, p<.05) and interaction
between groups and time (V=.204, F(5,59)=3.015, p<.05) was found. Univariate analysis shows that Sphericity is satisfied for EU performance (F(1,63)=12.096, p<.001, η2 =.161) and EU strengthening (F(1,63)=1.394, p<.05, η2=.093). These effects are very weak but suggest that the difference in scores between pretest and posttest are significant for the attitudes strengthening and performance. However, this analysis does not take any variance in groups in consideration which means that nothing can be said about the effect of deliberation (versus
13
non-deliberation) on these attitude dimensions. Finally, when looking at an interaction effect between time (pretest/posttest) and groups (deliberators/non-deliberators) a significant interaction effect was found for (F(1,63)=4.644, p<.05, η2=.564) the negative affection dimension. A profile plot has been drawn to look closer at the interaction effect (see figures A1, A2, A3, A4 and A5 in appendix). The results indicate that the experimental group, the deliberators, scored higher on the negative affection dimension after deliberating than the control group did.
DISCUSSION
As hypothesized, deliberation did have an effect on EU attitudes. While deliberation had an effect on EU attitudes, the effect did not vary across all five dimensions. Deliberators significantly increased in their negative affection towards the EU compared to the control group while there were no significant effects found on the other EU attitudinal dimensions. The increase of negative affection among deliberators is a complete opposite of the
deliberation effect during EuroPolis in 2009. While Sanders (2012) found that deliberators became more pro-European during EuroPolis, this experiment caused deliberators to increase negative affection.
A possible explanation for the difference between these results is the context of the experiment. Sanders (2012) already pointed out that the outcomes of EuroPolis might be caused by the style of the event and the vision of the organizers. Participants of the EuroPolis were invited over to Brussels and participated at a deliberative experiment in the buildings of the European Parliament. The setting of this deliberative experiment may be accountable for a large proportion of the positive attitude changes (Sanders, 2012). The present deliberative experiment was not held in Brussels and did not take place within the European Parliament.
14
Instead, it was conducted at the University of Amsterdam in a laboratory setting. Perhaps the context of the EuroPolis was the only reason for the positive effects on EU attitudes and the present results are the actual effects of deliberation on EU attitudes. This suggests that deliberation actually has a negative effect on EU attitudes.
Another explanation might be the time difference. Ritzen, Zimmerman and Wehner (2013) state that euroscepticism has increased, especially since the financial crises. A possible explanation for the different deliberation effects between the present study and EuroPolis in 2009 might be that public opinion has become more eurosceptical. A third possible
explanation for the increase in negative affect towards the EU among deliberators might lie in the theory of priming. Immigration was one of the topics discussed in the experiment and priming can activate a set of links in the cognitive network (Scheufele & IIyengar, 2012). Therefore, by deliberating about immigration, negative affections that were not salient before the experiment could have been activated during the experiment. This might have led to saliency of a more negative set of links regarding attitudes towards the EU and might have caused the increase in negative affection after deliberation.
Possible moderation effect can be found in issue salience. Fishkin, Luskin and Hahn (2007) found that the greater the salience of the subject discussed, the smaller the absolute net change on attitude. At the present time cost-benefit evaluations of EU membership might be salient due to the war between Ukraine and Russia in Eastern Europe and increased terrorist threat. Due to saliency of EU issues, the effect of deliberation on EU attitudes might have been too weak and therefore not significant.
While the deliberation effect on EU attitudes did not vary across all five dimensions, it did vary across some. The fact that deliberation did have a significant effect on negative affection while not affecting any of the other dimensions, confirms the multifaceted nature of EU
15
attitudes conceptualized by Boomgaarden et al (2011). If EU attitudes were a unidimensional concept, the present results were unexplicable. The present study and its findings are therefore in support of the claim made by Boomgaarden et al. (2011) that EU attitudes are of a
multifaceted nature and a multidimensional approach is necessary.
There are numerous limitations that need to be addressed. The convenience sample was not representative for the population. Diversity in terms of age, education and income were very low and groups did not represent a random sample of society. Another consequence of the convenience sample was the fact that group participants were not always strangers from each other. This could increase the probability of group participants deliberating in a way that is social desirable. Not only familiarity with other group participants might have led to social desirability. Participants were part of the social network of the moderator. This meant that participants were familiar with the moderator. Although the moderator was deliberately neutral and laid out rules for deliberation, social desirability might have played a role in participants’ effort in deliberation. The results also showed some clear questionnaire effects (see figure A1 in appendix). These might have occurred because of the short amount of time between the pretest and posttest questionnaires. Another limitation of this experiment was the small amount of participants (N=65). This limitation might have contributed for a large proportion to the limited amount of observed significant effects.
Although the present study has its limitations, the present study contributes knowledge to the interest of scholars and policy-makers. First, the study contributes to the empirical evidence that deliberation has an effect on EU attitudes. Second, in line with the conceptualization of EU attitudes by Boomgaarden et al. (2011) the significant effects of deliberation demonstrate a multidimensional concept of EU attitudes. Third, results of this deliberative experiment point to a misinterpretation of the deliberation effects during EuroPolis in 2009. The
16
pro-European attitudes, has been mistaken for the effects of impressive contextual factors. The results of this deliberative experiment in a laboratory setting suggest that the actual deliberation effect, without impressive contextual factors like the ambiance of a European Parliament, on EU attitudes is negative. Scholars should conduct future empirical research on deliberation and its effects on EU attitudes to confirm or falsify the present findings. It is recommended to vary contextual factors to exclude external influences that might be mistaken for effects of deliberation. Until proven wrong, this study implies that a deliberative
17
REFERENCES
Andersen, V. N., & Hansen, K. M. (2007). How deliberation makes better citizens: The Danish Deliberative Poll on the euro. European journal of political research, 46(4), 531-556.
Barabas, J. (2004). How deliberation affects policy opinions. American Political Science
Review, 98(04), 687-701.
Bernhagen, P., & Schmitt, H. (2014). Deliberation, political knowledge and vote choice: Results from an experiment with second-order elections. European Union Politics, 1465116514531506.
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.
Caluwaerts, D., & Reuchamps, M. (2014). Does Inter‐group Deliberation Foster Inter‐group Appreciation? Evidence from Two Experiments in Belgium. Politics,34(2), 101-115.
Chambers, S. (2003). Deliberative democratic theory. Annual review of political science, 6(1), 307-326.
Christin, T., & Trechsel, A. H. (2002). Joining the EU? Explaining public opinion in Switzerland. European Union Politics, 3(4), 415-443.
De Vries, C. E. (2013). Ambivalent Europeans? Public Support for European Integration in East and West. Government and Opposition, 48(03), 434-461.
De Vries, C. E., & Edwards, E. E. (2009). Taking europe to its extremes extremist parties and public Euroscepticism. Party Politics, 15(1), 5-28.
18
Dryzek, J. S. (2000). Deliberative democracy and beyond: liberals, critics, contestations. Oxford University Press.
Fishkin, J. S., & Luskin, R. C. (2005). Experimenting with a democratic ideal: Deliberative polling and public opinion. Acta Politica, 40(3), 284-298.
Fishkin, J. (2009). When the people speak: Deliberative democracy and public consultation. Oxford University Press.
Follesdal, A., & Hix, S. (2006). Why there is a democratic deficit in the EU: A response to Majone and Moravcsik. JCMS: Journal of Common Market Studies, 44(3), 533-562.
Gastil, J., Bacci, C., & Dollinger, M. (2010). Is Deliberation Neutral? Patterns of Attitude Change During" The Deliberative Polls™". Journal of Public Deliberation, 6(2), 3.
Gastil, J., Black, L., & Moscovitz, K. (2008). Ideology, attitude change, and deliberation in small face-to-face groups. Political Communication, 25(1), 23-46.
Habermas, J. (1990). Moral consciousness and communicative action. MIT press.
Himmelroos, S., & Christensen, H. S. (2014). Deliberation and Opinion Change: Evidence from a Deliberative Mini‐public in Finland. Scandinavian Political Studies, 37(1), 41-60. Hobolt, S. B., Spoon, J. J., & Tilley, J. (2009). A vote against Europe? Explaining defection at the 1999 and 2004 European Parliament elections.British journal of political science, 39(01), 93-115.
Isernia, P., & Fishkin, J. S. (2014). The EuroPolis deliberative poll. European Union
Politics, 15(3), 311-327.
Luskin, R. C., Fishkin, J. S., & Hahn, K. S. (2007, September). Deliberation and net attitude change. In ECPR general conference, Pisa, Italy (pp. 6-8).
19
Luskin, R. C., Fishkin, J. S., & Jowell, R. (2002). Considered opinions: Deliberative polling in Britain. British Journal of Political Science, 32(03), 455-487.
Mendelberg, T., & Oleske, J. (2000). Race and public deliberation. Political
Communication, 17(2), 169-191.
Muhlberger, P. (2005, September). Attitude change in face-to-face and online political deliberation: Conformity, information, or perspective taking. InAmerican Political Science
Association Annual Meeting, Washington, DC.
Mutz, D. C. (2008). Is deliberative democracy a falsifiable theory?. Annu. Rev. Polit. Sci., 11, 521-538.
Neblo, M. A., Esterling, K. M., Kennedy, R. P., Lazer, D. M., & Sokhey, A. E. (2010). Who wants to deliberate—and why?. American Political Science Review, 104(03), 566-583.
Sanders, D. (2012). The effects of deliberative polling in an EU-wide experiment: Five mechanisms in search of an explanation. British Journal of Political Science, 42(03), 617-640.
Scheufele, D. A., & Iyengar, S. (2012). The state of framing research: A call for new directions. The Oxford Handbook of Political Communication Theories. New York: Oxford
UniversityPress.
Stoeckel, F. (2013). Ambivalent or indifferent? Reconsidering the structure of EU public opinion. European Union Politics, 14(1), 23-45.
Thompson, D. F. (2008). Deliberative democratic theory and empirical political science. Annu. Rev. Polit. Sci., 11, 497-520.
20
Wojcieszak, M. (2011). Deliberation and attitude polarization. Journal of
Communication, 61(4), 596-617.
Zhang, W. (2012). Perceived Procedural Fairness in Deliberation: Predictors and Effects. Communication Research, 0093650212469544.
21
APPENDIX
Table A1. Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Gender 65 0 1 .51 .50 Age 65 1 4 1.6 .81 Education 65 2 8 6.63 1.63 Political interest 64 1 7 4.28 1.70 Political ideology 65 1 9 5.65 1.89 Pretest affective dimension 65 1 4.6 2.02 .91 Pretest identity dimension 65 1.2 6 3.41 1.19 Pretest performance dimension 65 1 6.2 3.68 1 Pretest utilitarianism dimension 65 1 7 4.78 1.18 Pretest strengthening dimension 65 1 7 2.75 1.11 Posttest affective dimension 65 1 4.4 1.94 .81 Posttest identity dimension 65 1 6 3.48 1.08 Posttest performance dimension 65 1.6 5.6 3.92 .88 Posttest utilitarianism dimension 65 1 7 4.91 1.11 Posttest strengthening dimension 65 1 6 2.96 1.02
22
Table A2. Levene’s Test of Equality of Error variances.
F Df1 Df2 Sig.
Pretest affective dimension ,781 1 63 ,380
Pretest identity dimension ,810 1 63 ,372
Pretest performance dimension ,301 1 63 ,585
Pretest utilitarianism dimension ,291 1 63 ,592
Pretest strengthening dimension 1,573 1 63 ,214
Posttest affective dimension ,395 1 63 ,532
Posttest identity dimension ,043 1 63 ,837
Posttest performance dimension ,115 1 63 ,735
Posttest utilitarianism dimension 1,651 1 63 ,204
Posttest strengthening dimension ,420 1 63 ,519
23
Table A3. Five dimensional concept of EU attitudes (Boomgaarden et al., 2011).
Multdimensional concept of
EU attitudes
Zeer mee oneens (1) (2) (3) (4) (5) (6) Zeer mee eens (7) Negative AffectionIk ben bang voor de Europese Unie. (1)
Ik voel me bedreigd door de Europese Unie (2)
Ik ben boos op de Europese Unie. (3)
Ik walg van de Europese Unie. (4)
De Europese Unie is een bedreiging voor de Nederlandse identiteit en cultuur. (5)
Identity
Burger van de Europese Unie zijn betekent veel voor mij. (6)
Europeanen delen een gemeenschappelijk erfgoed, cultuur en geschiedenis. (7)
Ik voel me verbonden met Europeanen. (8)
De Europese vlag betekent veel voor mij. (9)
Ik ben trots dat ik een Europese burger ben. (10) Performance
Het besluitvormingsproces van de EU is transparant. (11)
De EU functioneert goed. (12)
De EU verspilt veel belastinggeld. (13)
De EU functioneert volgens democratische principes. (14)
Ik vertrouw het Europees Parlement. (15)
Utilitarianism
De EU bevordert vrede en stabiliteit. (16)
De EU bevordert bescherming van milieu. (17)
Lidmaatschap van Nederland bij de EU is goed. (18)
Nederland profiteert van zijn lidmaatschap bij de EU. (19)
Ik profiteer van Nederlands' lidmaatschap bij de EU. (20)
Strengthening
De EU moet één land worden. (21)
Ik ben voor meer besluitvorming en beleid vanuit de EU. (22)
De invloed van de EU op besluitvorming moet vergroot worden. (23)
De EU moet snel uitbreiden. (24)
De EU moet uitgebreid worden met andere landen. (25)
24
Figure A1. MANOVA result indicating significant differences after deliberation.
25
Figure A3. MANOVA result on EU Performance, not significant
26
