1
Is Nationalism truly Independent?:
Thesis on the Mechanism through
which Nationalism affects
Euroscepticism
Name: Daan ValkStudent ID: 10338195 Teacher: Ph. D. Seiki Tanaka
Course: Varieties of democracy and democratization Study: Political Science
University of Amsterdam 4678 words
[Abstract]
This paper puts forth the argument that exclusive nationalism, a strong connection to the nation with an explicit absence of connection to the EU, becomes a strong indicator of Euroscepticism when the future of the nation is perceived to be in peril. Preliminary data seems to indicate otherwise however, and it is thus concluded that the two factors, though significantly correlated with Euroscepticism, are independent of each other.
2
[Introduction]
The European Union (EU) has taken center stage in recent years. The past two decades have been filled with ups and downs for the EU, from the introduction of the Euro to the outcome of the BREXIT-vote. This shows that the EU has become controversial. Euroscepticism, a negative attitude towards the EU and the further integration of it, has been on the rise and many populist political parties that have risen throughout Europe in this period are explicitly campaigning with Eurosceptic messages. National identity has been cited as a major banner of this countermovement. Yet not all people who consider national identity as important have turned against the EU.
In this paper we will explore how national identity is connected to Euroscepticism. By analyzing survey data we hope to find the mechanisms that cause people to view the nation as incompatible with the EU. The question we hope to answer is thus:
Why do some people who give great and exclusive importance to their national identity view the EU negatively, while others don’t?
Because the EU has become increasingly important through numerous reforms and crises, its future has become more relevant to the future of all Europeans. Understanding what drives the debates over its future has thus also become important. Though the connection between national identity and Euroscepticism is often made, clarifying the connection in a more fundamental way will contribute to the larger scientific debate.
3
[Literature review]
The term Euroscepticism, as it is commonly used, refers to the general aversion towards the EU and the further integration of the EU from either a political party or citizen’s perspective (van Klingeren, 2014: p13; Mudde, 2016: p77; Vasilopoulou, 2017: p23). Further research has opted to create subdivisions, as the term Euroscepticism is considered by many as a misnomer, too broad a term to meaningfully describe the issues involved (Duff, 2013: p140). Van Klingeren (2014: p13) offers differentiations between utilitarian vs affective Euroscepticism, based on either costs and benefits or emotions and ideals, and diffuse vs specific Euroscepticism, evaluation of the whole or of specific parts of the EU. Similarly, Mudde(2016: p77) also offers a differentiation between diffuse vs specific support for EU policy and integration, but he and researchers after him then go on to devise a
spectrum of Euroscepticism, some along a one-dimensional axis between pro and anti-EU, others adding more dimensions to it (Vasilopoulou, 2017: p23). These include Mudde’s identification of Euroenthusiasts, Europragmatists, Eurosceptics and Eurorejects, as well as rejectionist, revisionist, minimalist, gradualist, reformist and maximalist formulated by other authors (Vasilopoulou, 2017: p23).
This article will use the classification formulated by Boomgaarden et al. (2011), as it is focused specifically on individual attitudes rather than party positions and it was formulated based on survey data similar to the data used for this paper. This classification identifies five aspects of
Euroscepticism: affection, performance, identity, utilitarianism and strengthening. Affection refers to emotional attitudes towards the EU, such as fear or anger due to seeing the EU as some form of threat (Boomgaarden et al., 2011: p 258). Identity refers to the formation of a European identity or lack thereof. Performance entails the functioning of the EU, both economically as well as politically. Utilitarianism is related to cost/benefit analyses of EU membership on multiple levels. Lastly,
strengthening refers to attitudes towards the future development of the EU, including but not limited to further EU integration (Boomgaarden et al., 2011: p 258).
Although the literature shows that Euroscepticism is hard to define, generally speaking a consensus can be identified from these differing definitions about the aspects encompassing it. These are largely based on the suspected causes. Euroscepticism has been studied for several decades. Over time the presumed causes and consecutive research on those specific topics has changed. Originally, the cause of Euroscepticism was thought to be purely economical in nature, assuming that lower income households would suffer more from the increased competition between national groups within a further integrated EU common market would lead people with lower income to be against the EU (van Klingeren, 2014: p14-15). Similarly, nations who were already in economic problems would view greater EU involvement as further destabilizing to their economy (van Klingeren, 2014: p14-15). The root assumption of these theories was the rational choice theory, that human actors based their
4
attitudes on personal interest and an analysis of the direct consequences further integration would have on their livelihoods (van Klingeren, 2014: p14). As such, the same criticism often faced by rational choice-based theories applies here.
Chief amongst these was that the assumption that people always know or understand the EU when forming their attitudes, is incorrect (McLaren, 2002: p553). Research showed this not to be the case. To those that do not know or understand the EU, voting on matters pertaining to it, such as referendums or elections, was argued to be based on their evaluations of national governments (McLaren, 2002: p553). As a result, new research followed on the subject of perception and information (McLaren, 2002: p553)
Further research also investigated social factors such as cultural identity as a possible cause (McLaren, 2002: p552). It was argued that national identity was such an important factor that European integration, like the common market and freedom of movement, would cause conflict among groups based on their identities (McLaren, 2002: p552). People tend to evaluate policies and institutions not just on their own personal interests, but on those of their national community (McLaren, 2007: p236). The assumption that national identity can form a driving factor behind an individual’s political standpoints is based in part on the dichotomy between individual and group level value allocation, namely the theory of Societal interest values (Funk, 2000: p 38). This theory states that when people form their policy or political opinion, they can do so based on the impact a policy has on their society, even when it has no impact on them personally (Funk, 2000: p 38-39). It is important to take into account that this indicates people support their societal group out of more than just self-interest, but as part of their value-system.
It has been shown previously that people tend to mentally separate their experiences and perceptions of the public and private sphere’s (Funk, 2000: p 38-39). People have a tendency to be protective of any group that they identify themselves with (McLaren, 2007: p237; Hooghe & Marks, 2004: p 416). National communities are by many considered to be the primary identity group they associate with, leading to the assumption that people will be very protective of those communities specifically (McLaren, 2007: p237).Furthermore, EU integration is considered part of a larger trend, including globalization, of erosion of the nation-state (McLaren, 2002: p552). People who strongly associate with their nation are thus more likely to perceive EU integration as a threat (van Klingeren, 2014: p15).
One important caveat here is that there is a significant difference between exclusive and inclusive national identity (Hooghe & Marks, 2004: p 416). People usually identify with more than one group or community. Even if they have the strongest connection with their national identity group, this does not exclude identification with other groups, such as regions, cities and also Europe. Those people who have an inclusive sense of national identity are often found to be more, not less,
5
supportive of the EU. Only exclusive national identity, when one only associates oneself with one identity group, has a strong correlation with Euroscepticism (Hooghe & Marks, 2004: p 416).
Partly as an extension of the nativist argument above, Immigration to and within the EU was also considered to be an important factor of determining Euroscepticism (De Vreese & Boomgaarden, 2005: p64). Research has shown some people having a tendency to pay more attention towards groups than others, classifying in-groups as their own groups and out-groups as the ‘other’. Out-groups tend to be looked on unfavorably compared to in-groups and people who categorize immigrants as out-groups tend to categorize people into out-out-groups in general. As such, people who negatively look on immigration are more likely to also look negatively on EU integration. As the EU has many
nationalities and ethnicities within its borders, they are less likely to view other European nationalities as part of their in-group, including the European identity as a distinct group (De Vreese &
Boomgaarden, 2005: p64).
In-groups and Out-groups are sometimes perceived to be at odds with each other, competing for the same resources (van Klingeren, 2014: p30). These can be physical resources, such as jobs or housing, but also cultural resources, such as values or political power (van Klingeren, 2014: p30). As such, the growth or rise to prominence of one group, either in number of people or presence in the public discourse, can be viewed as a detriment to other groups (van Klingeren, 2014: p30).
6
[Argument]
It has thus been established that a strong and exclusive sense of national identity is an important factor in the formation of Eurosceptic political opinion. Yet, we can’t say that all people who identify exclusively with their national group are by definition also anti-EU. There must be other factors at play, that lead people to place higher priority on their national group or that cause the perception that their group is at odds with the EU. As not all people agree on this incompatibility, as shown by those people that identify with both (and often even more than two (Hooghe & Marks, 2004: p 416)), we venture to identify the circumstances that cause exclusive national identity to become an important political motivation.
This thesis seeks to find the mechanism that causes nativist attitudes to also become anti-EU attitudes. I argue that nativist attitudes only become politically driving when it is coupled with a sense of fear or harm to the national community. As political attitudes and voting behavior can be
significantly pivoted on national concerns, such an assumption is quite possible. It can even be called somewhat of an open door, as obviously nationalistic people respond to matters pertaining to the nation. Although this connection is not a given, as there are other reasons than just a strong connection to the national community that can spur nationalistic views. Harm to the group can also result in harm to its members, so individualistic motivations can also be at play.
As stated in the introduction, the question this thesis seeks to answer is the following:
Why do some people who give great and exclusive importance to their national identity view the EU negatively, while others don’t?
This paper hypothesizes that people who look more negatively to the future of their national community, and that don’t have an established connection with Europe, will be more likely to feel threatened by the EU than people who are not negatively inclined to the future of the nation. The mechanism assumed in this hypothesis is that the national wellbeing only becomes an important issue to people, when it is perceived to be declining. In such a situation people will look for a cause outside of the group, as group dynamics shows often occurs, and people that do not feel connected with the EU will likely be moved against it, as group related topics are abundant when it comes to the EU, including dynamics between the many nations it is comprised of, integration into a European group and external groups when it comes to immigration and refugees. Thus we expect to observe that people who are pessimistic about the future of their nation, will be more inclined to look negatively on the EU in general.
An alternative hypothesis, as already mentioned, is that people are not particularly concerned with the wellbeing of the nation, but that they are worried about their own future or that of their household. They merely view the nation as an extension of themselves, that a deterioration of their
7
own situation is caused by the deterioration of the national situation. If that is the case, we expect to see people being negatively inclined towards the EU, even if the national situation is not deteriorating but their own situation is.
8
[Method]
The argument and subsequent hypothesizes stated in the previous chapter will be tested through the use of multiple linear regression. An existing dataset will be used, that of the Eurobarometer survey conducted in November 2017 (European Commission, 2017). From the dataset, the Netherlands specifically will be used as a case. The choice was made to focus on a particular country as case, rather than using the entire dataset, as doing so would require adding several factors of note to the
calculations, such as differences in the overall national situations, cultural differences such as political left-right dynamics or historical differences such as relatively recent regime changes. Though doing so would improve the validity of the outcome, it was not possible at the time.
The Netherlands was chosen primarily for two reasons. The first being that it is one of the founding members of the EU, a member of the Eurozone yet has seen a significant rise in Eurosceptic attitudes in the past two decades. As such, it can be presumed that a disconnect with the EU is not the result of the country only having joined it recently. The second reason is that the research of
Boomgaarden et al., on which the definition of Euroscepticism that is used here, was also done with the Netherlands as case.
As mentioned, Eurobarometer was chosen as dataset for this thesis. This was done with two considerations in mind. First, Eurobarometer has comprehensive data about the EU specifically. Questions like knowledge about the EU and whether someone identifies with the EU are amongst the variables measured, making it ideally suited for answering the research question. Secondly,
Eurobarometer conducts surveys annually, meaning the data is likely to be as recent and thus relevant as can be found, short of conducting a survey personally.
The data provided by Eurobarometer will be analyzed via the statistics analysis program SPSS, and will primarily consist of multiple linear regressions. This method is chosen as it is the most suited to analyze how multiple independent variables affect a dependent variable. Specifically, the regression will be used to test the existence of significant correlations, the measure of overlap between variables and the degree to which the independent variables explain the dependent variables.
9
Variables:
The dependent variable used for this paper is Euroscepticism. It is measured by combining the data of four questions into one variable. As the surveys were conducted in Dutch, all questions
henceforth paraphrased will be translations. These questions are:
• “Would you say it is going in the right direction or in the wrong direction in the EU?” • “What image does the EU in general evoke for you, from very positive to very negative?” • “Are you satisfied with how democracy functions in the EU, from very satisfied to not at all?” • “To what degree do you agree with the statement: The Netherlands would have a better future
outside the EU?”
These four variables have been recoded into dichotomous variables, so that a significantly negative attitude towards the EU is coded as 1 and the other answers as 0. These variables are then combined into a single ordinal variable, simply called Euroscepticism, with 5 values from 0 to 4. The higher the number, the more Eurosceptic the correspondent is.
Next, three independent variables will be used for the regression, the first of which is
exclusive nationalism. This indicates a high degree of attachment to the national identity, but explicitly not with the EU. This is measured by combining the data of two questions from the dataset. These questions are:
• “To what degree do you value the Netherlands?”
• “Do you view yourself as: only Dutch, Dutch first but also European, European first but also Dutch or only European?”
These questions were recoded, such that only the answers confirming a singular national identity were coded as 1 and the rest as zero. Then they were combined and recoded again by making 1 correspond with both questions being answered in an exclusive nationalistic way. The resulting dichotomous variable is then used in the regression.
Negative future outlook of the nation is the second independent variable. Three questions were chosen from the data set to give a general indication of how someone looks towards the future of their national community. These questions were: “What are your expectations for the coming year, will things be better, worse or the same when pertaining to:
• The general situation in the Netherlands? • The economic situation in the Netherlands? • The job market in the Netherlands?”
10
They were recoded into dichotomous variables with 1 being worse and 0 being better or the same. Those were then combined into a single variable from 0 to 3, indicating to what degree a person looks negatively towards the future of their nation.
The third independent variable is very similar to the previous one, except that it is based on question about a person’s outlook on their personal situation, rather than the nation. This variable was introduced to test the alternative hypothesis stipulated previously. The questions were almost identical as well. “What are your expectations for the coming year, will things be better, worse or the same when pertaining to:
• Your life in general?
• The financial situation of your household? • Your work situation?”
These were then similarly recoded into dichotomous variables, which were then also combined into a single variable ranging from 0 to 3, indicating the degree to which someone is worried about their personal future or that of their household.
Lastly, a total of four control variables are used. The first two of which are age and gender. Gender was recoded into a dichotomous variable by recoding 0 to mean male and 1 to mean female. Age was already recoded in the dataset as an interval variable divided into 5 age groups of ten years, starting at 15-24 and one group of 65 years and older. These control variables were introduced as basic tests to ensure correlations are not caused by age or gender gaps.
The third control variable gives an indication of political ideology through a left-right placement scale. People were asked to indicate where they saw themselves on the political spectrum by placing themselves on a scale of 1 to 10, 1 being left and 10 being right. This has been recoded into two dummy variables, one combining the three most left indicating answers and the other combining the three most right indicating answers. These indicate a strong left-leaning and a strong right-leaning ideological position respectively. Political ideology has been added to test whether any of the
independent variables are strongly overlapping with it. If this does turn out to be the case, it will be hard to say whether one’s ideology is affected by nationalism or future outlook, or whether ideology affects them, but we will reflect on that later, if it turns out to be the case.
The fourth control variable used here is EU knowledge. This is indicated by testing people’s knowledge about three statements about the EU. In the survey correspondents were asked to say whether they thought the statements were true or false. These statements were:
• “The Eurozone currently consist of 19 member states.” (true)
• “Members of the European Parliament are directly chosen by the citizens of every member state.” (true)
11
• “Switzerland is a member of the EU.” (false)
The answers to each question were recoded into dichotomous variables, with 1 indicating a correct answer and 0 indicating a wrong answer or when correspondents indicated they didn’t know. These were then combined into a single variable, ranging from 0 to 3, where 3 indicates only correct answers being given. This variable was included to test whether knowledge about the EU affects someone’s likelihood of seeing it as a threat. The questions used don’t say much about a person’s understanding of the EU, how much power it has or what its intentions are, so it is not considered a perfect indicator, but it gives some indication of basic understanding.
12
[Data-analysis]
In this chapter the regressions analysis takes place, as well as describing the steps taken to get the results. As this thesis specifically uses the Netherlands as a case, while the dataset used contains data from multiple countries, all results are sorted between cases that indicate a Dutch nationality and those that don’t. Only the results from those separated cases are then used.
First, we run a basic linear regression to see whether there is a correlation between the main dependent variable, Euroscepticism, and independent variable, exclusive nationalism. As can be seen in table 2, the results show a coefficient of 0.600, meaning that people who have an exclusive
nationalist view are on average 0.6 points higher on the scale of Euroscepticism, which has been coded as a scale that runs from 0 to 4, with 4 meaning very Eurosceptic and 0 not at all. The alpha value, indicating the level of significance, is lower than 0.001, indicating the odds of this result being an accidental outlier is less than 1%. Lastly, the adjusted R2 indicates how much of the dependent variable is explained by the independent variable. The indicated value of 0.044 means 4.4% of Euroscepticism in this model can be explained through exclusive nationalism.
After this, the second independent variable, negative national outlook, is tested. Model 2 shows it is again significant with an alpha value lower than 0.001. The constant also remains roughly the same. Two things are of note when comparing with the previous regression. The unstandardized coefficients are similar, but the one from model 2 is smaller, indicating a smaller effect from negative national outlook. This can however not be said exclusive nationalism is a dichotomous variable whereas negative national outlook is ordinal, meaning a step of 1 in either cases means different things. The same is true for the standard deviations. The second thing of note is that the R2 is more twice as high in model 2, at 0.107. This indicates that the model can explain the Euroscepticism variable for 10.7%. Presumably this means that negative national outlook has a larger effect on Euroscepticism than exclusive nationalism does.
Next the first multiple linear regression is performed to test the main argument of this thesis, whether the effects of exclusive nationalism on Euroscepticism can be explained by negative national outlook. Model 3 shows the results. The model retains a high degree of significance, while the unstandardized coefficients remain relatively the same, only slightly smaller. The R2 value 0.146 indicates an even higher degree of explanatory power from this model, which is not necessarily what was expected. The last value of note the degree of collinearity. Identified through the tolerance value, it indicates the degree of overlap between multiple independent variables. Tolerance values are not shown in the table, but for this model it was 0.998. The higher the tolerance value, between 0 and 1, the less the amount of overlap between variables.
13
This very high tolerance value, coupled with the fact the R2 value is almost exactly the same as adding the previous two models together and the coefficients barely changed, indicates virtually no overlap between these two independent variables. If our hypothesis were correct, and the effect of exclusive nationalism is, at least in part, triggered by the onset of negative national outlook, there should have been a significant amount of overlap between the two variables. This not being the case would indicate our hypothesis must therefore be rejected.
However, to ensure the validity of the models used so far, several control variables are introduced. Models 4 through 8 test the main model by adding the aforementioned control variables separately. Model 9 than adds all of them. We won’t go through all of them in detail, but only discuss the results of note. The two models introducing ideology seem to indicate that people who view themselves as rightwing are more inclined to be Eurosceptic, whereas people who view themselves as leftwing are less inclined. The degree of significance is much lower for both correlations however, so whether it is reliable is questionable. Gender is coded as 0 meaning the correspondent is male and 1 female, so the negative coefficient of gender indicates men are more likely to be Eurosceptic than women, although here too the significance is relatively low. The final model shows little effect on the main independent variables, and even in this model, the tolerance values do not drop below a value of 0.9, so no significant overlap can be found.
Lastly, to test the alternative hypothesis, models 10 through 12 include the intendent variable for negative personal outlook. Model 10, a singular linear regression, tests the effect of this variable on Euroscepticism. When compared to model 2, the effects are smaller across the board. When tested with exclusive nationalism, the effects are similarly smaller than in model 3. Only the tolerance value is slightly lower, at 0.993 compared to 0.998. Though this is relatively small, it does seems to indicate a stronger connection between a negative personal outlook and exclusive nationalism than between negative national outlook and exclusive nationalism. The final model adds both negative outlook variables, show the highest effect yet, though only on each other and the tolerance value still does not drop below 0.9.
14
[Conclusion and reflection]
To conclude this thesis, the results of the previous data analysis will be discussed here and the hypotheses stipulated previously will be considered. After that, the author will reflect on the process and results of the paper and recommend improvements and future research.
The results of the regression analysis indicates that the two independent variables, exclusive nationalism and negative national outlook, are not influenced by each other in a significant enough way to accept the hypothesis put forth in this thesis. Both seem to have a significant effect on
Euroscepticism, as previous research indicates, even after the control variables are taken into account. The alternative hypothesis yielded similar results and must therefore also be rejected, though it also yielded a significant correlation with the dependent variable.
Reflecting on this research, I must admit there are an abundance of possible improvements to the models used in this paper. As discussed in the literature review, Euroscepticism is a broad concept that is best measured as separate dependent variables, rather than adding it’s parts together as was done here. Furthermore, the dependent variables of negative outlooks both indicate mostly economic anxiety, neglecting the factor of security and culture to a large degree. These factors were hard to incorporate due to a lack of good indicators in the Eurobarometer dataset, yet deserves to be mentioned nonetheless.
As for the control variables, it was intended to include both income and education as a control variable, however the dataset did not contain a satisfactory indicator for either. It had one that was intended as an indicator of education level, but it was far from ideal so including it did not seem like a meaningful addition. Media consumption was also planned as a control variable, however it would have required adding over ten separate recoded variables which I was not able to do in time unfortunately.
As for why unstandardized coefficients were used for the analysis, I honestly don’t exactly remember why but I was told not to use those in assignments form previous classes so opted to not do so here again. I don’t believe it made much of a difference for the analysis.
15
[References]
Boomgaarden, H. Schuck, A. Elenbaas, M. de Vreese, D. (2011) Mapping EU attitudes: conceptual and empirical dimensions of Euroscepticism and EU support, European Union Politics 12:2, pp 241– 266.
De Vreese, C. H., & Boomgaarden. H. G. (2005). Projecting EU referendums: Fear of immigration and support for European integration. European Union Politics, 6, pp 59-82.
Duff, A. (2013) On Dealing with Euroscepticism. Journal of Common Market Studies. 51:1, pp. 140-152
European Commission (2017) Eurobarometer 88.3, GESIS, Cologne: ZA6928, dataset version 1.0.0, Source: dx.doi.org/10.4232/1.13007, last visited 19-6-2018
Fanoulis, E. (2017) Knowledge of the EU and citizen participation in European governance: an agonistic democracy perspective, European Politics and Society, 19:1, pp 35-48
Funk, C. L. (2000) The Dual Influence of Self-Interest and Societal Interest in Public Opinion,
Political Research Quarterly, 53:1, pp 37-62
Hooghe, L. Marks, G. (2004) Does Identity or Economic Rationality Drive Public Opinion on European Integration? PS, Political Science and Politics 37:3, pp 415–420
van Klingeren, M. (2014). "Welcome" to Europe: How media and immigration affect increasing
Euroscepticism, Ipskam Drukkers: Enschede
McLaren, L. M. (2002). Public support for the European Union: Cost/benefit analysis or perceived cultural threat? The Journal of Politics, 64, pp 551-566.
McLaren, L. M. (2007). Explaining mass-level Euroscepticism: Identity, interests, and institutional distrust. Acta Politica, 42, pp 233–251
Mudde, C. (2016) On extremism and democracy in Europe. The EU, Euroscepticism and Brexit, Routledge
Vasilopoulou, S. (2017) ,Theory, Concepts and Research Design in the Study of Euroscepticism from:
16
Appendix
Table 1Name variable N % missing
(absolute) Mean Std. deviation Minimum Maximum Euroscepticism 1054 0% (0) 1.268 1.230 0 4 Exclusive Nationalism 1054 0% (0) 0.253 0.435 0 1 Outlook Nation 1054 0% (0) 0.330 0.749 0 3 Outlook Personal 1054 0% (0) 0.202 0.541 0 3 Age 1054 0% (0) 55.59 15.173 15 87 Gender (1=♀) 1054 0% (0) 0.498 0.500 0 1 Left-Right placement 1032 2,09% (22) 4.98 1.879 1 10 Knowledge EU 1054 0% (0) 1.753 0.804 0 3
17
Table 2
Model 1 Model 2 Model 3 Model 4 Model 5
Exclusive Nationalism .600***
.561***
.536*** .542***
(0.085)
(.080)
(.081) (.080) Outlook Nation
0.538*** .524***
.522*** .533***
(.048)
(.047)
(.047) (.047) Ideology right
.272*
(.129) Ideology left
-.204* (.083) Knowledge EU Age Gender
Constante
1.116***
1.090***
.953***
.938*** 1.002***
(.043)
(.039)
(.043)
(.043) (.047) N
1054
1054
1054
1054 1054
R2 adjusted 0,044 0.107 0.146 0.149 0.150 Note: unstandardized coefficients with Std. Error in brackets
18
Table 3
Model 6 Model 7 Model 8 Model 9
Exclusive Nationalism .577***
.560***
.571***
.553***
(.080)
(.080)
(.080)
(.081) Outlook Nation
.515***
.511*** .531***
.516***
(.047)
(.047)
(.047)
(.047) Ideology right
.170
(.130) Ideology left
-.171* (.083) Knowledge EU .123*** .113** (.044) (.043) Age .077*** .078*** (.024) (.024) Gender -.156* -.143*
(.070)
(.070) Constante
.736***
.604***
1.025***
.500***
(.088)
(.116)
(.054)
(.143) N
1054
1054
1054
1054
R2 adjusted 0,152 0.153 0.149 0.166 Note: unstandardized coefficients with Std. Error in brackets
19
Table 4
Model 10 Model 11 Model 12
Excl. Nationalism .563*** .545*** (.084) (.080) Outlook Person .406*** .369*** .184** (.069) (.068) (.067) Outlook Nation .487*** (.048) Ideology right Ideology left Knowledge EU Age Gender Constante 1.186*** 1.050*** .931*** (.040) (.044) (.043) N 1054 1054 1054 R2 adjusted 0,031 0.070 .151
Note: unstandardized coefficients with Std. Error in
brackets *** p < 0,001, ** p < 0,01, * p < 0,05
