Individual Euroscepticism in Hungary:
Drivers and trends, 2002-2016
University of Amsterdam
2019
Graduate School of Social Sciences
Political Science MA
European Politics and External Relations
Thesis supervisor: dr. Armen Hakhverdian
Second reader: dr. Eelco Harteveld
Name: Bálint Halász
Student number: 12296066
Submission date: 24 June 2019
1 Table of contents:
1. Literature review
1.1. Euroscepticism in general
1.2. Sources of individual Euroscepticism 1.2.1. Views on the Economy 1.2.2. Institutional (dis)Trust
1.2.3. Cultural Incentives and National Identity 1.2.4. Immigration
1.2.5. Additional Factors
1.2/6 Academic Explanation of the Choice of the Dependent Variables 1.3. Euroscepticism around Europe
1.4. East-Central European Euroscepticism 1.5. Euroscepticism in Hungary
2. Data and methodology 2.1 Data and variables 2.2. Dependent variables 2.3. Independent variables 2.4. Control variables 2.5. Methodology 3. Analysis and discussion
3.1. Trust in the European Parliament amongst Hungarians (2002-2016) 3.2. View on the state of European Unification (2004-2016)
3.3 Selected independent variables separately
3.3.1. Public perception of the effects of immigration on the country (2002-2016) 3.3.2. Hungarian attitudes towards gay rights (2002-2016)
3.4 Regression-analysis
3.4.1 Regression model 1: Trust in the European Parliament 3.4.2. Comparison and trends
3.5 Regression model 2: European unification 3.5.1 Trends and comparison
3.6. Comparing the two type of models 4. Concluding Remarks
2 List of figures:
Tables:
1. Standardized ß coefficients in model 1. (Trust in the EP) 2. Standardized ß coefficients in model 2. (European Unification) Graphs:
1. Mean score of trust in the European Parliament in Hungary (2002-2016)
2. Mean score of feelings about the state of European unification in Hungary (2004-2016) 3. Mean score of immigration's perceived effect on Hungary (2002-2016)
4. Proportion of Hungarians with a negative stance on gay-rights (2002-2016)
5. Trends in standardized ß coefficients of model 1. (Trust in the European Parliament) 6. Trends in standardized ß coefficients of model 2. (European Unification)
3 Introduction:
The refugee crisis of the mid-2010s reshaped Hungarian political debate completely. With an enormous governmental hate campaign, Viktor Orbán and his party Fidesz managed to create a one-dimensional public debate where everything was about migration. The boundaries of this campaign, however, has not stopped with issues regarding the flow of refugees into Europe as it soon became connected to something even bigger, the European Union. Slogans like “Stop Brussels” or “Brussels wants to settle illegal migrants into Hungary” started echoing in the country, helped by the gigantically swollen government friendly traditional and online media. Billboards with similar slogans on them flooded every street, wall, city-light and advertising pillar. The plan of so-called relocation quotas proposed by the European Commission in 2015 soon became the number one target of such permanent campaign and governmental communication, which resulted in a highly controversial referendum in 2016, where the electorate could decide whether they “want the European Union to be able to mandate the obligatory resettlement of non-Hungarian citizens into Hungary even without the approval of the National Assembly?”. Even though the referendum had no result as less than 50% of eligible voters turned up, the proportion of those who voted “NO” was an astonishing 98%. This meant that the government convinced nearly 3,5 million Hungarians, that, —to put it plainly— the European Union wanted to make them live with people, who constituted a danger to their lifestyle, Christian heritage and culture. But how has this previously overwhelmingly pro-EU society turned onto Brussels during the years of the refugee crisis and ever since? What is the nature of Hungarian individual Euroscepticism? What are the drivers behind anti-EU feelings in the country? Can it be just about migration as it seems in recent years? Or is it more a cultural-religious thing, which the government is so eager to protect the country from the bureaucrats of Brussels? Are economic grievances the most important factors in Euroscepticism? Could it be the case of institutional trust? Are they all important?
Hungarian Euroscepticism was looked at mainly from the direction of party politics. Party-based Euroscepticism is mostly limited to the right side of the political spectrum (Dúró, 2014). The ruling party’s negative stance on the EU is unquestionable and is described as a great example of both governmental (Dúró 2016) and soft-Euroscepticism (Taggart & Szczerbiak, 2004). Individual Euroscepticism is unfortunately relatively under-researched in the country and the region with only a narrow scope of findings, which show a case somewhat similar to what the general literature describes (Závecz, 2011).
Thus, my research question is the following: What kind of factors drive and what is the nature of Hungarian individual Euroscepticism? In order to find an answer to this question conducted a multivariate regression analysis using data from the European Social Survey (ESS). I used four types of factors that are mentioned in the existing literature as sources of public Euroscepticism and applied them to the Hungarian case. There are previous researches on this topic in the Hungarian context, however to my knowledge neither of them uses a longer timeframe. Therefore, my analysis aims to fill this gap by
4
concentrating on not just one round of the survey, but eight of them from 2002 to 2016. By this, I was able to draw conclusions regarding extended trends in the impact of certain factors of Hungarian Euroscepticism and managed to connect some of them to political changes and events. My main findings show a relatively stable picture regarding the impact of the four types of factors I examined. In the case of the Trust in the European Parliament institutional distrust was the biggest contributor to Eurosceptic sentiments and the case of the views on European Unification anti-immigrant sentiments and cultural incentives proved to be the most impactful. In terms of possible political reasons behind the trends in these variables’ effect on Euroscepticism, I managed to connect the growing importance of immigration sentiments and cultural reasons to the Orbán Government’s Eurosceptic and radical right-wing turn.
In the first section, I review the existing literature on the topic. First, I look at general Euroscepticism literature, which will be followed by a more thorough depiction of individual/public Euroscepticism. After it, I look at works about Euroscepticism around Europe, with a particular focus on Central Eastern Europe. In the last subsection of the first part, I concentrate on literature regarding specifics of Hungarian anti-EU sentiments. In the second section, I present my data-gathering and methodology. The third and biggest section contains the results of my research divided into four parts. Firstly, I provide a brief overview of the dependent variables and two of my independent variables. Secondly, I present the results of the regression analysis using my first dependent variable. This is followed by the third subsection with the results regarding my second dependent variable. This section is concluded by a comparison between the results of the separate regression analyses. In the fourth and final section, I conclude my findings and present ideas and suggestions for future research.
1. Literature review:
1.1. Euroscepticism in general:
Euroscepticism is an always evolving phenomenon with a multi-faceted nature. A constantly changing beast with no general definition. The term “Eurosceptic” was first used by journalists in the 1980’s United Kingdom as a label for certain Members of Parliament who were to some extent against the newly joined European Integration (Leruth et al, 2018).
The phrase has strong shortcomings in the academic world. First, in the first decades, it has not been an academically defined term, which lacked a certain ideological core. Therefore, it was not concrete whether Euroscepticism is a new phenomenon or just another label for something else, like nativism or populism. Secondly, it is a negative construction from the beginning, which means it can be anything that is anti-EU. Thirdly it has temporal and geographical specialties, thus Euroscepticism cannot be dealt with as a whole accurately (Leruth et al, 2018). However, we can turn to several definitions and descriptions. According to the Oxford Dictionary Eurosceptic means “A person who is opposed to increasing the powers of the European Union”. In the Encyclopaedia Britannica’s interpretation, Euroscepticism is a “European political doctrine that advocates disengagement from the European
5
Union (EU). Political parties that espouse a Eurosceptic viewpoint tend to be broadly populist and generally support tighter immigration controls in addition to the dismantling or streamlining of the EU bureaucratic structure” (Ray, 2016). And for example, Taggart and Szerbiak conceived Euroscepticism as “an encompassing a range of critical positions on European integration, as well as outright opposition.” (Hooeghe & Marks, 2007, p. 2.)
The above authors also distinguished between two types of (party-based) Euroscepticism. “Hard Eurosceptics” are those, who are against European integration as a whole and want their respective country either not to join or leave the EU. On the other hand, “soft Eurosceptics” are only opposing certain parts and policies of the European Union (Taggart & Szczerbiak, 2004). The literature about Euroscepticism can be divided into two equally important segments, party-based and individual (public/popular) Euroscepticism. As my thesis focuses on the latter one, in the next section I provide an overview of the literature on this phenomenon.
1.2. Sources of individual Euroscepticism:
To sum up the literature on my topic, it can be said that individual Euroscepticism is mainly based on four types of factors. The first one regards economic concerns, nevertheless the pieces I present differ on whether these negative feelings are based on personal utilitarianism or general economic sentiments. The second group of factors regards general distrust in political institutions. The third one is based on cultural and identity issues, which contains a wide range of elements from religion to exclusive national identity. The last one is anti-immigration-based Euroscepticism. My research is structured accordingly as I chose four independent variables that could cover each said possible source of anti-EU sentiments.
1.2.1. Views on the Economy:
The literature on economically driven Euroscepticism can be differentiated into two parts. The first one concerns the driver of anti-EU feelings is a general dissatisfaction with the economy. The second one is based on personal utilitarianism, where the Eurosceptic sentiments are linked to the citizen’s personal economic interests.
Looking at the relationship between economical attributes and support of the European integration, Gabel and Whitten pointed out that the public support of the EU is mainly based on economic criteria. They argue that people are more concerned about the “subjective” rather than the “objective” economy. They found that as economies change, the support for the EU fluctuates with it (1997). In comparison to previous researches (Gabel & Palmer, 1995) they state that the citizens care about the economy itself, rather than the benefits provided by the EU, therefore, more beneficial policy by Brussels won’t affect the level of Euroscepticism (Gabel & Whitten, 1997). These leads us to the second aspect of Euroscepticism based on economic views.
In terms of personal utilitarianism-based Euroscepticism Lauren McLaren found in her work on mass level Euroscepticism, that personal interest-based utilitarianism to be one of the strongest sources of
6
Eurosceptic sentiments (2011). Matthew Gabel also drew the same conclusion, as he stated, that utilitarian appraisals of integrative policies are not very important predictors of Eurosceptic feelings in every member state of the EU (1998). Another shortcoming of the economy-based Euroscepticism was found by Van Elsas, Hakhverdian and van der Brug, as they found evidence regarding the impact of economic issues in Euroscepticism, however only amongst left-wing citizens (2016).
1.2.2. Institutional (dis)Trust:
Another source of individual Euroscepticism according to the literature is institutional distrust amongst citizens. An important distinction has to be made amongst national and supranational institutions in this case in terms of the effect they respectively cause. McLaren argues, that it’s the feelings about institutions beyond the national level that bear a stronger impact on anti-EU feelings. A negative view of national institutions in her view, however, can work as a proxy for dissatisfaction with supranational ones, or in other cases, these two factors can act together and strengthen each other as some sort of general dissatisfaction with the entire political system (2011).
Other researches, however, contain evidence regarding the bigger impact of national institutions. Boomgaarden et al. found evidence, that national government approval is one of the two factors that had a significant effect on the five EU attitude dimensions (emotional responses; sense of European identity; performance and democratic and financial functioning; utilitarian attitudes and future strengthening of the EU with further integration) they created (2011). People’s feeling that their governments can’t protect them from insecurity is also an aspect of how dissatisfaction with national institutions play a crucial part in Eurosceptic sentiments (Hooghe & Marks, 2007).
Matthew Gabel also supported this statement but, he only found it accurate in new member states of the European Union, as government approval doesn’t play an important role in anti-EU sentiments in the original member states (2011). When looking at Euroscepticism on the individual, national and regional level Lubbers and Scheepers also found that political distrust is an important element of negative feeling towards the Union (2007).
Another aspect to look at regarding the connection between feelings about institutions and feelings about the EU was presented by Harteveld et al. They investigated the accuracy of three theories regarding the trust in the EU, the logic of rationality, the logic of identity and the logic of extrapolation. They found the latter to be the best predictor of feelings about the European Union, as they argue that citizens appraise the whole integration project through the lenses of their own nation-states’ current circumstances. Therefore, the trust in European Institutions either formed by preexisting feelings regarding citizen’s home-state or both are originated in the same place (Harteveld et al, 2013). These results found by them are also indicators of the important part played by institutional distrust in Eurosceptic sentiments.
7 1.2.3. Cultural Incentives and National Identity:
The third element that can play a part in Euroscepticism around the continent is based on cultural and identity factors. Exclusive national identity amongst citizens can be a strong predictor of Eurosceptic feelings. In an article by Carey and Lebo, the authors found strong evidence regarding the substantially strong impact of identity issues in the formation of Eurosceptic sentiments. They also state that such an effect is even stronger than the previously discussed economic drivers (2001).
McLaren in her above-discussed article also supports the idea of the heightened importance of the fear of losing national identity amongst Europeans. In addition, she also argues, that these sentiments can strengthen the dissatisfaction with supranational institutions, therefore can play a part in political trust-based Euroscepticism (2011). In addition, not just an existing exclusive national identity can boost Euroscepticism, but the lack of a continent-wide European identity also (Hooghe & Marks, 2007). Van Elsas, Hakhverdian and van der Brug, when comparing anti-EU feelings on the right and left extremes of the political spectrum found, that both groups’ Euroscepticism is to some extent driven by cultural issues. However, on the right, these cultural and national-identity factors are the sole base of negative feelings towards the EU and are paired with a complete rejection of the European project. On the other hand, left-wing citizens’ Eurosceptic sentiments are also some part sourced in cultural issues, but not solely, as they are also concerned with economic problems (2016). Harteveld et al. also tested the impact of identity on the trust in the EU although, they found this theory less accurate compared to the other two they examined (see above) (2013).
1.2.4. Immigration:
The fourth element of individual Euroscepticism that is frequently mentioned in the relevant literature is anti-immigration sentiments. Opposing immigration is possibly the strongest connection link between Eurosceptic parties across Europe. The case is the same on the individual level as well as there is evidence regarding how the negative feelings towards immigrants can impact how people think about the EU.
Boomgaarden et al. found that the five attitude dimensions regarding the EU they differentiated are all strongly impacted by immigration-sentiments (2011). Lubbers’ and Scheepers’ article also support this statement, as they found that perceived threat from immigrants is a strong contributor to Eurosceptic sentiments (2007). It is enough to look at the British case and Brexit to take a hint on the impact which anti-immigration bears on anti-Europe feelings (Hooghe & Marks, 2007; Goodwin & Milazzo, 2015).
1.2.5. Additional Factors:
Apart from the four main types of drivers of individual Euroscepticism, the literature on the topic contains other sources that are worth mentioning. The impact of media framing is often mentioned in
8
the literature as a booster of Eurosceptic feelings (Hooghe & Marks, 2007). According to the article from the previously discussed Lubbers and Scheepers, time and space can also play a part in the level of public Euroscepticism, as they found that both the distance from Brussels and the duration of the EU membership can contribute to negative feelings towards the Union (2007). Socio-economic factors can also strengthen Euroscepticism on the individual level (Carey & Lebo, 2001). In addition, Theresa Kuhn discovered a connection between individual transnationalism and Euroscepticism. As she states, those who are more pro-transnationalism individually are less prone to be Eurosceptic and this is more the case in more globalized countries (2011).
1.2.6 Academic Explanation of the Choice of the Dependent Variables:
I chose the two dependent variables because of their good explanatory nature regarding two different aspects of feelings about the European Union. By using trust in the European Parliament, I managed to look at, how people think about the institutional side of the EU and also investigate how citizens view the state of democracy in the Union. The EU is a “political union in the making”, therefore the trust in it, can show how people think about the process itself and how “healthy” the whole polity is (Camisao, 2015 p.1). Where the trust or distrust in the EU originates was examined by Harteveld et al. They tested three theories of possible logic behind such trust and found that in this case, the logic of extrapolation was the most accurate. That logic means, that the citizens of the EU project their trust in their respective national institutions on to the EU itself (2013).
On the other hand, the question of the state of the European Unification covers a different aspect of EU-sentiments. Rejecting further EU integration (i.e saying that the unification has already gone too far) can be an example of hard Euroscepticism (Taggart & Szczerbiak, 2004) and also right-wing Euroscepticism (van Elsas et al, 2016). Even though one would think that saying the unification process should go further means that the respondent has a positive view on the EU, supporting further integration is also a point of critique by mostly left-wing citizens with Eurosceptic views, who think that the way the Union functions is insufficient. (van Elsas et al, 2016)
1.3. Euroscepticism around Europe:
In this section, I provide a brief review of the literature on how Euroscepticism looks like in different regions and states around the European Union. Condruz-Bacescu differentiates four types of Euroscepticism that vary across the continent: Euroscepticism based on (1) economic-, (2) sovereignty-, (3) democratic- and (4) political criterion (2004). Given the most prominent example of the impact of Euroscepticism, Brexit, I start with the United Kingdom. According to Evans and Butt, British anti-EU feelings, in general, are less driven by economic issues and are more about the exclusive national identity (Hooeghe & Marks, 2007). Before the referendum on the membership was held Goodwin and Milazzo, in their piece on what drives Euroscepticism, found that in the British case the pro-Brexit groups’ Euroscepticism is based on two equally important factors, anti-immigrant sentiments and the alleged
9 democratic deficit of the EU (2015).
When it comes to Scandinavia —which is considered to be the least Eurosceptic part of the Union — Raunio pointed out the differences between Denmark, Sweden, and Finland. According to him these differences on the extent of Euroscepticism lies in the way the countries party system and party preferences look like, and the way their respective governments are formatted (Hooghe & Marks, 2007). On the other hand, there is Southern Europe, which has very strong Eurosceptic sentiments amongst the general public. Llamazares and Gramacho found on the topic, that the southern Euroscepticism has two pillars, cultural exclusivism, and issues regarding economic redistribution (Hooghe & Marks, 2007). After the economic and euro-crises—which had the strongest impact on this region— anti-EU feelings were strengthened by the sentiment that the Brussels leadership forced states like Spain and Greece to accept things they did not want, in order to avoid the collapse of their economies. Fears amongst the public linked to unemployment, economic insecurity led to a certain type of Euroscepticism which can be described as opposition to a “foreign intervention” by the Northern member states (Condruz-Basescu, 2014, p. 6.).
1.4. East-Central European Euroscepticism:
After discussing three other regions of the EU, I turn to the most important region for my research, East-Central Europe. In general, it can be said, that this region’s countries, a decade after they joined the European Union became worryingly filled with Eurosceptic forces. Countries like Poland or Hungary are currently under huge pressure from EU institutions because of alleged violations of basic values of the integration. This institutional pressure is embodied in the so-called Article 7. procedure, which, should it have gone through, would result in the respective countries loss of vote in the European Council (POLITICO, 2018). There is very limited literature on individual Euroscepticism in the region, therefore I concentrate on the works regarding the party-based side, which is also important in reaching a broader understanding of how anti-EU sentiments work in these countries.
Neumayer looked at the politicization of EU issues in three new member states of this region, Poland, Hungary and the Czech Republic from the 1990s (2008). She stated that there has been a widespread political consensus amongst parties of the region on the importance of EU accession, however also on certain critiques of EU policies. According to her, mentioning and using the theme of Europe has been an instrument in reshaping the political spectrum on three levels. First in differentiating between mainstream parties (pro-EU) and protest or outsider parties. Secondly mainstream political actors from their competition. And finally, certain current inside parties from each other. In general, the pro-EU stance has become a political norm and parties only colluded in a limited fashion on it, in order to present themselves to the European leadership as legitimate actors. European issues were widely used by parties in order to boost their support and gain political capital. On a pragmatic level, saying “yes, but” to the EU has become the consensus in party competition (Neumayer, 2008).
10
Czech Republic, and Hungary), Dúró found a growing tendency in mainstream (mostly right-wing) parties criticizing the EU. The Polish Law and Justice Party (PiS) and the Hungarian Fidesz Civic Alliance are the two most prominent examples of Eurosceptic political actors, nonetheless, the Czech ANO and some Slovak parties can also be considered growingly anti-EU. The first two are also sole governmental parties, and the Czech party is also in the government coalition leading the country (2016).
1.5. Euroscepticism in Hungary:
After I looked at the academic works on the broader region, I concentrate on the existing literature on Hungarian Euroscepticism. After the EU accession in 2004 a rapid, but not substantially powerful change occurred in the public opinion regarding the Union, as a the long-hoped major increase in Hungarian living standards didn’t come true. This disappointment was also connected to a previous similar one, which happened after the regime change in 1989, where the majority of Hungarians thought that “western standards of living” were close, however, the gap between the two regions remained the same. Given this, a decline in support of the EU was visible, even though, not as big as in the neighboring countries (Kész, 2014).
As Molnár points out in her piece on a comparison between Italian and Hungarian Euroscepticism, the economic crisis of 2008 strengthened the already existing effect of the above-explained “post-accession crisis” (2016). The devastating effects of the recession further empowered a feeling of being semi-peripheric and led to a growing number of Eurosceptic and populist parties like Jobbik. Molnár describes it as an “EUphoria” that turned into “EUphobia” (Molnár, 2016, p. 20).
If we look at the party-based Euroscepticism in the country the thing that differentiates Hungary from nearly all member states of the EU, is the fact that the Viktor Orbán led Fidesz-KDNP government is considered to be a strongly Eurosceptic party. Even though they are not against the country’s membership in the EU, they are loud and harsh critics of certain EU policies and attributes. There is an ongoing anti-EU campaign in Hungary with many points where they attack the Union and the Brussels leadership. An interesting paradox relationship can be observed regarding the Hungarian governments EU policies. Despite the strong critical stance on core EU values and policies, the Western Balkan enlargement is strongly supported by the Orbán government, even though this can also be explained by national interest incentives. Given this, Fidesz’s Euroscepticism can be considered to be a part of the “soft” side of the Taggart-Szczerbiak terminology (2004). If we look at other major parties, we can find two more, to some extent, Eurosceptic one: Jobbik, which used to be an extreme right party with “hard Eurosceptic” positions, recent changes in the party’s ideology makes them a soft-Eurosceptic party. On the other hand, Politics Can Be Different (LMP) criticizes the EU from the left, with strong anti-big capital messages. All other major opposition parties are pro-EU (Kész, 2014).
When looking at the Hungarian individual Euroscepticism Gergő Závecz checked five types of individual variables on their effect of Eurosceptic sentiments, using Eurobarometer 73.4 data from 2010 (2011). These groups of independent variables were the following: (1) perceived economic changes, (2)
11
exclusive national identity, (3) trust in national government (4) party preferences, cognitive political mobilization, and political values, and (5) demographic variables. The selection of such variables was in line with the literature on individual Euroscepticism I presented above. He found that nearly all identified attributes had an effect on Euroscepticism in the country, with a few exceptions, that can be a result of country-specific circumstances (Závecz, 2011).
2. Data and methodology:
2.1 Data and variables:
My analysis is based on data from eight rounds of European Social Surveys, from 2002 to 2016 (europeansocialsurvey.org, 2019). I used data from every ESS conducted in Hungary in order to present a long timeframe through which drivers of Hungarian individual Euroscepticism can be thoroughly described. Using the SPSS tables of each round of surveys I decided to do a multivariate regression analysis with two dependent, four independent and nine control variables. The data was collected during six consecutive weeks between September of the year of the survey and January of the following year by ESS, using face-to-face (CAPI) interviews, with a sample size above 1500 in each case (europeansocialsurvey.org, 2019) In the next subsection, I discuss every variable I used, after the name of each, I write the shortened version of it in brackets. These shortened names can be seen in my tables and graphs in the analysis section.
2.2. Dependent variables:
The two dependent variables were the Trust in the European Parliament (‘Trust in the EP’) and the European unification gone too far/ should go further (‘European Unification’). Every respondent could answer these questions asked in the survey on a scale from zero to ten. In the case of the first one, 0 means no trust in the EP and 10 means complete trust. The question on the European unification works the same way, as 0 means the unification of European countries has already gone too far, whereas 10 means it should go further. Unfortunately, this second variable wasn’t asked in 2002 and 2010 in the surveys. Even though the two variables capture different approaches of the public opinion on the EU, both are accurate ways to measure how the respondents feel about certain aspects of the Union. In addition, it must be stated, that the first dependent variable can be a source of flaws in the results, as the question about the respondents’ trust in the EP is asked right after a series of questions regarding other types of institutional trust.
2.3. Independent variables:
I choose four independent variables (1) Immigration makes the country better or worse to live in (‘Immigration’), which is also measured on a 0 to 10 scale, where 0 means immigration makes living in the country worse and 10 means it makes it better. This variable was chosen in accordance with both the general literature on Euroscepticism – which states, that one of the main reasons why citizens feel
12
negatively of European integration is immigration –, and because the Hungarian government’s recent anti-immigration campaigns, through which the Viktor Orbán led cabinet criticizes the European leadership, Brussels and altogether the whole EU. The second independent variable was (2) Gays and lesbians are free to live as they wish (from now on ‘Stance on Gay Rights’), which captures the cultural-religious aspect of anti-EU feelings that can be found in the literature on the topic. This variable is measured on a 1 to 5 scale where 1 is strong agreement and 5 is strong disagreement with the statement. The third variable I choose on the independent side was a Political Trust Index, which I created from three separate variables of the survey: Trust in the country’s parliament, Trust in Politicians and Trust in political parties. These three correlate very highly in every round of the survey, are measured on the same 0 to 10 scale, therefore I decided to look at their average score, which I created by combining their score in every round and I divided the result by three. My fourth and final independent variable was an economic one, The government should reduce income inequalities (‘Economic views’). This variant is in line with the Euroscepticism literature’s economic motivation theory and captures the fourth major aspect of sources of anti-Europe sentiments. This variable is also measured on a 1 to 5 scale where 1 means the respondent strongly agrees, whereas, 5 means they strongly disagree.
2.4. Control variables:
As control variables, I used nine separate ones. The first one is Gender, which is followed by Age, the Highest level of education, Household total net income (‘Household income’), Belonging to a certain religion or denomination (‘Belonging to a Religion’), and How religious are you (‘Level of Religiousness’). Unfortunately, the variable regarding income was excluded in the first three rounds of ESS. Additionally, for further accuracy from the variable regarding educational levels, I created four dummy variables, which were the following: (1) Secondary Education, (2) Upper secondary education, (3) Postsecondary education without tertiary education (‘Postsecondary Education’) and (4) Postsecondary education with tertiary education (‘Tertiary Education’). The reference variable was the lowest possible educational level that could be chosen by the respondents, Lower than secondary education.
2.5. Methodology:
After identifying these variables, I decided to do a multivariate regression analysis in order to measure how each separate variable effect the public’s view on the European Union. I used the SPSS software to conduct such regression analysis by putting first the Trust in the EP variable in the dependent side and all the others into the independent one. After it, I did the same with the other variable, European Unification. I ran the regression analysis on every round of the ESS survey and identified the Standardized ß coefficients and p values. It was necessary to use the standardized version of the Beta, as some of the variables are measured on different scales, therefore using the unstandardized one would have led to further flawed results. By this, I produced two regression models. By identifying every
13
coefficient, I was able to compare every independent variable’s individual effect on the two dependent ones in every survey-year. This comparison enabled me to create graphs and tables where the level of effect on the above two dependent variables, and the trends of such effect is clearly visible throughout the timeframe of 14 years. I supplemented every table with two additional values: (1) Number of cases (N), and (2) R Square. P-values appeared behind the ß coefficients in the form of stars, with the number of it based on their value. One star meant, that the p is lower than 0,05, two stars were an indicator of a p-value under 0,01 and three stars meant that in that case, the p is smaller than 0,001.
3. Analysis and discussion:
3.1. Trust in the European Parliament amongst Hungarians (2002-2016):
In the first section of my analysis, I briefly discuss the two dependent variables separately. If we look at the Trust in the European Parliament in Hungary through the 14 years between 2002 and 2016 (Graph 1.), we can observe a small decline in the mean score given by respondents. In 2002, before the accession, the mean level of trust amongst was 5,67. until 2010 a constant decrease can be seen, with the lowest score in 2008, which was only 4,12. Such a decline can be explained by a number of reasons. As I already discussed in the literature review, after Hungary joined the European Union in 2004, the public image of the EU suffered a minor blow, which was caused by the so-called “post-accession crisis”. The global economic crisis had a heavy impact on the country, and as the very low 2008 trust-level shows, the feelings toward the EU suffered from it deeply. After 2008 a strong increase is visible, as the level of trust went up by more than 0,6. But, in the next two years, the mean score went down again and got close to the negative record of the year when the global recession hit the country. The year 2014 is extremely important regarding EU-Hungary relations, as that was the last year before a strong anti-EU campaign by the Orbán government took off. The results of it can be seen in the decline of trust between 2014 and 2016. However, given the intensity and scope of the anti-EU campaign by the government, with billboards with “Stop Brussels” on them, and even a referendum against the EU refugee-relocation quotas, the decline is relatively small.
Overall it can be said, that the level of Hungarian trust in the EP is relatively stable, with only weak fluctuations. It is also clearly visible, that until 2016, the governmental anti-EU campaign hadn’t had a substantially big effect on the public image of the EP, although, this may have changed in recent years.
14
Graph 1. Source: European Social Survey (2002-2016) 3.2. View on the state of European Unification (2004-2016):
Just as the Trust in the European Parliament shows a relatively stable picture. Even though we are missing data from the 2002 and 2010 rounds of ESS, a very similar pattern can be seen (for visibility purposes I included the trend line to this graph), with the highest mean score of 5,52 from 2004 and a slightly lower on in the last round with 4,12. The overall decline is 1,3, which shows that as the years have gone by, Hungarians became less supportive of further European unification. The above mentioned governmental campaign targeting the European leadership had an effect on how the public views the future of integration, as from 2014 to 2016 the mean score dropped by more than 0,4. This decrease is roughly similar to the previous case and shows that the campaign may have had an effect, however not as big as the size and intensity would have projected. Other effects may have been in play however, the so-called post-accession crisis’s impact is also visible in the dropping mean score in the first two available rounds (2004, 2006). On the other hand, the year of the recession had no impact on how people view the European integration process, as from 2006 even a slight increase occurred.
Graph 2. Source: European Social Survey (2004-2016) 0 1 2 3 4 5 6 2002 2004 2006 2008 2010 2012 2014 2016 Me an Sco re Year
Mean score of trust in the European
Parliament in Hungary (2002-2016)
0 1 2 3 4 5 6 2002 2004 2006 2008 2010 2012 2014 2016 Me an Sco re YearMean score of feelings about the state of
European unification in Hungary (2004-2016)
15 3.3 Selected independent variables separately:
In this section, I provide a brief overview of the trends of two of my independent variables, Immigration, and Stance on Gay Rights. These two are in my opinion crucially important to look at separately as trends in these can be indicators of general trends in Hungarian public view, which is very important in the later regression-analysis.
3.3.1. Public perception of the effects of immigration on the country (2002-2016):
Even though immigration to Hungary has always been very low and conjured up mostly by ethnic Hungarians moving to the country from neighboring countries like Romania, Slovakia, Ukraine and Serbia, the topic of migrants and migration is by far the hottest one in recent years of public debate. Since the 1990’s public view on immigration was fluctuating strongly around certain events like the flow of Yugoslavian refugees (Számado, 1992) in the early years of the last decade of the 20th century or the debate about the status of ethnic Hungarians in the early 2000s (Weinstein, 2004).
It can be seen that the mean score of whether Immigration makes the country worse or better place to live has never been high, as it never went above 4,4, which means that the Hungarian public is mostly against immigration. In 2002, and 2004 the mean score is the same (4,02), yet a drop can be seen in 2006 data, which could be a result of a highly controversial referendum about dual citizenship for ethnic Hungarians in neighboring countries. This referendum sparked a very xenophobic campaign from political parties like the Alliance of Free Democrats (SZDSZ) and the Hungarian Socialist Party (MSZP), with previously used slogans like granting dual citizenship means “20 million Romanians will come to the country” (Weinstein, 2004). This highly anti-immigrant slogans first appeared during the first Orbán government’s decision to grant certain additional rights to foreign ethnic Hungarians, which was called the “status-law” (Deets, 2008). This is a great example of how xenophobic sentiments can be found on both sides of the political spectrum.
Graph 3. Source: European Social Survey (2002-2016) 0 1 2 3 4 5 2002 2004 2006 2008 2010 2012 2014 2016 Me an Sco re Year
Mean score of immigration's percieved effect
on Hungary (2002-2016)
16
The last important change occurred during the refugee crisis, which peaked in 2015 when hundreds of thousands of mostly Middle Eastern refugees marched through Hungary to reach their preferred western destination. During this time the third Orbán government (2014-2018) started to use immigration as their most important campaign topic. The government stance on immigration was very hard and negative as billboards with anti-immigrant slogans flooded the country (Thorp, 2015), a fence was built on the Serbian border to stop immigrants from entering Hungary and a substantially controversial referendum was held on the so-called “relocation quotas”. Even though this referendum resulted invalid as less than 50% of the electorate voted, the result was described by the government as “politically valid” because 98% of the those who participated voted against these quotas (Adler, 2016). The level of xenophobic sentiments skyrocketed. As the Tárki Institute found in their annual Xenophobia Index- which differentiates three groups regarding their position on immigration (anti-immigrants, undecided and immigration-friendly)- that by 2016, 99 % of the respondents were either anti-immigration or undecided (old.tarki.hu, 2016). The ESS results show a similar (even though, less extreme) picture as the mean score went down by 0,5. This also means that the crisis and the governmental campaign around it had some impact on how citizens view immigration.
3.3.2. Hungarian attitudes towards gay rights (2002-2016):
If we look at the trends on how Hungarians stand on issues regarding homosexuality we can see that the percentage of respondents who disagree or strongly disagree with the statement, that gays and lesbians should live freely as they wish was relatively stable from 2002 until 2014, with a proportion around 30%. However, after 2014, we can observe a huge increase in the percentage of those who to some extent oppose gay rights, as the 2014 26,1% jumped to 39,9 % in two years.
Graph 4. Source: European Social Survey (2002-2016)
No major event occurred regarding gay-rights, however, the government put a stronger emphasis on traditional values like family and religion –so in one word became more conservative— which could have resulted in a spike in homophobic sentiments. It is important to point out that same-sex marriage is still illegal in Hungary but, certain forms of partner-statuses are supported by law.
0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 2002 2004 2006 2008 2010 2012 2014 2016 Perc en ta ge Year
Proportion of Hungarians with a negative
stance on gay-rights (2002-2016)
17 3.4 Regression-analysis:
In this section, I present the results of my multivariate regression analysis. As I mentioned in the methodology and data section, I conducted a regression analysis with two dependent, four independent and nine control variables in order to present a picture about sources of Euroscepticism in Hungary over fourteen years, from 2002 to 2016, using eight rounds of European Social Survey data.
3.4.1 Regression model 1: Trust in the European Parliament:
In this subsection, I present my findings regarding the standardized ß coefficient in my regression analysis between the independent and control variables and the Trust in the EP. In the table below (Table 1.) there is every coefficient of the regression between the dependent variable and the independent and control ones and the lowest two lines contain the number of respondents (N) and the R Square. The level of p values is also represented in forms of stars behind some coefficients. There are some empty boxes in the table as there were some cases where no data were available to form the ESS database, however, the number of such cases is relatively low, therefore have no meaningful effect on my analysis.
Standardized ß coefficients in model 1. (Trust in the EP)
2002 2004 2006 2008 2010 2012 2014 2016
Immigration 0,213 *** 0,066* 0,083** 0,101** 0,101*** 0,035 0,041 0,186*** Stance on Gay Rights -0,087** -0,043 -0,01 -0,099*** -0,032 -0,066** -0,115*** -0,088** Political Trust Index 0,565*** 0,53*** 0,529*** 0,591*** 0,51*** 0,438*** 0,472***
Economic Views 0,01 0,024 -0,081** 0,004 -0,013 0,074** 0,02 -0,043
Gender 0,022 -0,032 -0,025 -0,023 0,047 0,008 0,018 0,028
Age -0,078 -0,003 -0,021 -0,058* -0,009 -0,082** 0,063* -0,016
Secondary Education 0,009 0,046 0,041 -0,051 -0,025 -0,042 0,102 -0,176** Upper Secondary Education 0,068 0,173** 0,138 -0,038 -0,026 -0,082 0,118 -0,288**
Postsecondary Education 0,142*** 0,051 -0,017 -0,028 0 0,016 -0,091 Tertiary Education 0,07 0,183*** 0,114* -0,02 -0,02 -0,048 0,106 -0,126 Household income 0,038 0,043 -0,003 0,145*** -0,013 Belonging to a religion -0,033 -0,001 -0,041 0,003 0,007 0,037 -0,082* 0,117** Level of religiousness 0,043 0,012 -0,001 0,075 -0,05 0,139*** -0,052 -0,034 R Square 0,076 0,369 0,322 0,353 0,378 0,351 0,278 0,335 N 1070 1089 1000 872 936 1102 979 846
Table 1. Source: European Social Survey (2002-2016) *p<0,05; **p<0,01; ***p<0,001 (two-sided).
The first independent variable (Immigration) has a fluctuating effect in terms of strength on the dependent one as the standardized ß varies from 0,035 to 0,213. The highest ß coefficient is from the first round and the lowest was recorded in 2012. The biggest decrease can be seen from 2002 to 2004, with 1,4, whereas, the biggest increase happened between 2014 and 2016. The ß is positive in every case, therefore the direction of the connection between the variables is similar, in other words, if one variable goes up, the other follows. Overall, it can be said that the strength of the ß is not very high although, in a survey-analysis cannot be considered insignificant either. The second independent
18
variable’s effect, which concerned views on gay rights varies from -0,01 to -0,115, with the lowest in 2006 and the highest in 2014. Every round produced negative ß-s therefore, the direction of the connection is antagonistic. In terms of strength, the same situation applies as in the case of the first independent variable. The third variable (Political Trust Index) has had a significant effect on the Trust in the European Union every year, with standardized Betas varying from 0,483 to 0,565. The highest beta was produced by the 2004 model and the weakest connection occurred in 2014. The connection between the independent and the dependent variable is unidirectional in every case, as every ß coefficient is positive. It must again be pointed out, that the results regarding the Political Trust Index can’t be considered completely accurate as a result of the structure of the ESS questionnaire. In the case of the last independent variable (Economic Views), the standardized Betas are fluctuating between 0,001 (2002) and -0,081 (2006). As it can be seen, the direction of the connection was not similar in every case as there are three instances (2006, 2010, 2016) when the Betas were negative. If we look at the coefficients, we can see that there are no very strong connections between the independent variable and the dependent one, however, because of the strong fluctuation, it is not true for every year. I will look at the trends and present a comparison between the ß coefficients in the next section. In terms of the control variables, we can see that the ß barely goes above 0,1 or under -0,1 and the directions of the connections are inconsistent. There were five cases where—as a result of lack of data— no ß could have been produced. The N varies from 872 to 1087 and the R Square was between 0,076 to 0,378, which means the independent variables explain the variation of the dependent variable in 7 to 37 percent of the cases. This proportion was the highest in my 2010 model, whereas it was the lowest in the case of the 2002 model.
3.4.2. Comparison and trends:
In this subsection, I discuss the trends of the ß coefficients and present a comparison between the effect of the four independent variables I was concentrating on (Graph 5.). It can easily be seen that the weakest explanatory variable was the fourth one, which regarded economic positions of the respondents. The ß coefficient provided by this independent variable fluctuates around zero without a consistent direction. It means that not just the strength of the impact of this variable is irrelevant, but a consistent direction of this weak connection is impossible to provide also. On the other hand, the effect of the Political Trust Index was the highest, in the territory around 0,5.
19
Graph 5. Source: European Social Survey (2002-2016)
The other two independent variables were fluctuating in the territory around 0,1, however, with different signs as the one regarding gay rights were in the negatives and the one regarding immigration in the positive territory. If we look at trends, we can see that all four variable’s impact on the Trust in the EP is quite stable with no strong fluctuations. The only variable which behaved in a relatively different way was the one concerning immigration. In the case of that, we can see two spikes one in the beginning and one in the end. Both spikes can be explained by political events, where immigration became a hot topic in the public debate. In 2002, the heated argument regarding the previously mentioned so-called status-law could have had an impact on how people connect the EU and immigration, and in the years between 2014 and 2016 where the Hungarian government initiated a very strong anti-EU campaign, where the EU was (and still is) criticized because of the mainstream immigration policy. It is fairly possible that the strength of the connection between the view on the EU and immigration became more powerful as a result of the continuous anti-EU and anti-immigration campaign of the Orbán Governments.
3.5 Regression model 2: European unification:
In this subsection, I present my findings of the regression analysis, with the dependent variable, European Unification go further/ gone too far (European Unification). The table below (Table 2.) contains every standardized ß coefficients I was able to produce during my research. Compared to the previous table with the results regarding the Trust in the EP, it is clearly visible, that the cases where no coefficients could have been calculated are significantly higher. These gaps are the result of lacking data, as for example in 2002 and 2010 when the public opinion on the dependent variable wasn’t measured by the European Social Survey. Even though this makes this part of the analysis less compact, I firmly believe that general trends and basic assumptions can be discovered also, just like in the case of the previous part.
-0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 ß CO EF FICI EN TS YEAR
Trends in standardized ß coefficients of model 1.
(Trust in the European Parliament)
20 Standardized ß coefficients in model 2. (European Unification) 2004 2006 2008 2010 2012 2014 2016 Immigration -0,049*** 0,142*** 0,116** 0,254*** 0,106** 0,267*** Stance on Gay Rights -0,092** -0,049 -0,111** -0,084** -0,164*** -0,181*** Political Trust Index 0,11*** 0,161*** 0,102** 0,048 -0,067* -0,062
Economic Views -0,009 0,043 -0,028 0,098** 0,102** 0,072*
Gender -0,021 -0,053 0,039 -0,005 0,02 0,011
Age -0,003 -0,013 0,024 -0,053 -0,089* -0,056
Secondary Education 0,058 0,068 -0,021 -0,104 -0,069 -0,136 Upper Secondary Education 0,11 0,076 -0,018 -0,213* -0,076 -0,286** Postsecondary Education 0,034 0,041 -0,016 -0,059 -0,009 -0,093 Tertiary Education 0,095 0,041 -0,021 -0,096 -0,076 -0,168* Household income -0,046 0,007 0,047 -0,037 Belonging to a religion 0,016 0,047 0,09* 0,138*** -0,015 0,069 Level of Religiousness 0,011 0,052 0,037 0,078* 0,042 0,127** R Square 0,064 0,073 0,058 0,152 0,073 0,188 N 1081 967 834 1059 940 812
Table 2. Source: European Social Survey (2004-2016) *p<0,05; **p<0,01; ***p<0,001 (two-sided).
If we look at the first independent variables effect on the public view on the depth of European integration, we can see that sentiments regarding immigration had a variously strong impact, with Beta coefficients from -0,49 to 0,267. The weakest and strongest impact is visible on the two ends of the timeline, with the 2004 model producing the lowest ß and the 2016 one the highest. The direction of the connection is not unanimous as there is one negative (2004) and five positive coefficients. The second variable which (Stance on Gay Rights) also had a fluctuating impact on the dependent variable with ß-s between -0,049 (2006) and -0,181 (2016). The direction of the connections in this case, however, was consistently antagonistic, with all of them being in the negative territory. The Political Trust Index had a varying impact on the views on European unification. The weakest effect was recorded in 2004 (0,48) and the strongest in 2006 (0,161). In terms of the direction of the connection between the independent and dependent variables, we can observe, that with being both negative and positive standardized ß-s- the course of the impact is inconsistent. The last independent variable (Economic Views) also had an irrelevant and inconsistent impact on the dependent variable with ß-s fluctuating between -0,028 and 0,102, and both negative (2x) and positive (4x) signs. The majority examined control variables’ effect varied between 0 and (+-)1, with only a few exceptions. The direction of such connections also showed an inconsistent picture with roughly equal proportion on the positive and negative part. The number of cases (N) was between 812 and 1081. The values of the R Square were around 0,1, with the highest in the last model (2016) with 0,188.
3.5.1 Trends and comparison:
A comparison between the variables impact on the views on European unification requires a distinction between the models before 2010 and after it (Graph 6.). In the first era, the strongest predictor of feelings about the EU was the Political Trust Index (0,102-0,161), closely followed by immigration sentiments (-0,049-0,116) and views on gay rights (-0,049--0,111). Economic Views finished last with ß-s very
21
close to 0, therefore, without any meaningful impact. After the break caused by the lack of data, the order of the most impactful variables changed. The effect of Immigration became the strongest predictor of feelings regarding the EU, with Betas between 0,106 (2014) and 0,267 (2016). Stance on Gay Rights turned into a strong runner up with ß-s between -0,084 (2012) and -0,181 (2016). The economic factor arrived third with ß-s around 0,1. Political trust seemingly completely lost its impact on EU sentiments with ß-s barely stronger than -0,05.
Graph 6. Source: European Social Survey (2004-2016)
These changes in the order predict interesting trends in the effect of separate variables. For instance, feelings about immigration went through an increase of more than 0,21 in absolute value, which means despite its relatively low impact in 2004 it became the strongest predictor of views on European unification amongst the four variables. Just in the case of the previous dependent variable, a huge jump is visible after the refugee crisis had hit Europe and the Hungarian government started its vicious campaign against the EU, because of its allegedly catastrophic immigration policies. In 2012, the standardized ß was only 0,106, but it increased by 0,161 to 0,267. Stance on Gay Rights was overall a more stable predictor compared to Immigration, although, a similar increase in power was observable after the 2010s. The ß of -0,084 more than doubled until 2016 (-0,181). This increase in the strength of impact can also be a result of the Hungarian government’s growingly conservative stance, which is also part of the EU-criticism as they frequently speak about the loss of Christian values and culture. If we look at the impact caused by political trust we can see a decline in its strength and also a change in the direction of such effect, as being one of the strongest variables with ß-s above 0,1 turned into values around -0,05 by the 2010s. Economic feelings have never had a strong impact on the view of European Unification with ß-s around 0 throughout this 12 years period. Overall regarding the 4 variables I examined, it can be said that except the economic one, slightly stronger fluctuations were visible compared to the first model.
-0.3 -0.2 -0.1 0 0.1 0.2 0.3 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 ß CO EF FICI EN TS YEAR
Trends in standardized ß coefficients of model 2.
(European Unification)
22 3.6. Comparing the two type of models:
In the previous subsections, I presented my findings regarding my four independent variables’ level of impact on the Trust in the EP and European Unification from the early 2000s to 2016. Now I turn to a comparison between these two types of regression models, with a particular focus on the order between the strength of the impact of these variables, and the trends amongst them in these two cases. In the first type of model, the most impactful variable on the trust in the EP was the political trust index by far throughout the years. On the other hand, there is no single strongest predictor of feelings about European unification as views on immigration and homosexuality were similarly strong sources. However, it must be also stated that the impact of political trust in the confidence in the EP is by far the strongest predictor I found throughout my analysis, nevertheless of the fact – as I mentioned in the methodology section— that it may be have been caused by flaws in the surveys. According to my findings, the independent variables with the least impact on the dependent ones were the economic factors in both cases, with standardized ß-s in the territory around 0. In terms of stability in trends, it can be said that the four variables’ effect on the trust in the EP was more constant with less and smaller spikes and downfalls. On the other hand, in the case of the model regarding European unification slightly more and bigger increases and decreases were visible, most importantly in the case of immigration-sentiments. It must be stated however, that despite the fact, that the Orbán governments had (has) a strong and fierce anti-EU campaign on the basis of immigration and traditional values and it resulted in some changes in the impact of my first two independent variables, the extent of these were significantly smaller than I expected when I started my analysis. To sum it up, despite some differences, the sources of Hungarian Euroscepticism examined by me has been relatively stable over time, and the Eurosceptic turn of the Hungarian government hadn’t had a really strong impact on them.
4. Concluding Remarks:
My analysis aimed to provide a thorough overview of drivers behind Hungarian individual Euroscepticism. Using the existing literature on individual Euroscepticism supplemented by country-specific factors, I identified four factors that might contribute to negative feelings toward European integration. I translated these four drivers into four independent variables using the European Social Survey. As for dependent variables I used two separate questions in the same surveys to cover two aspects of feelings toward the EU. After this, I conducted a multivariate regression analysis with said independent and dependent variables and additional nine control ones throughout 8 rounds of surveys from 2002 till 2016. In the analysis section, I presented my findings separately in the cases of the two dependent variables, which was followed by a comparison of the two. Given all this, I make four conclusions.
Firstly, compared to the existing literature, I found evidence to support that cultural incentives and institutional trust has an impact on the level of Hungarian Euroscepticism, however not for the same
23
extent when using the two different dependent variables. In the case of Trust in the European Union, institutional trust proved to be the strongest factor in Eurosceptic feelings, however, because of the structure of the questionnaire, such high values may have been compromised. Nevertheless, the same trust factors in the other case proved to be significantly weaker. In the case of the cultural-religious factor, I also found evidence for their impact, however, the other way around. Views on European unification were more affected by these drivers than in the case of Trust in the EP. My second conclusion is regarding another type of factor, the economic one. On the contrary to the literature, I barely found evidence to support, that economic incentives had an effect on Eurosceptic sentiments amongst Hungarians. Neither of the dependent variables were strongly impacted by the independent variable regarding economic components and in the case of European unification, it literally had no impact at all.
My third conclusion is regarding the influence of immigration sentiments. The immigration variable I examined had an impact on both dependent variables. Views on European unification was more influenced by feelings about immigration, than the trust in the European Union but, the strength of this impact was only substantial on the ends of my timeframe, in the early 2000s and in 2016.
My fourth conclusion regards the trends I discovered. Overall it can be said, that the level of the effect these factors posed on Euroscepticism was relatively stable throughout these years. Significant changes in Hungarian politics, like general elections or Viktor Orbán’s populist-Eurosceptic turn, seemed to have a limited impact on the trends. Although, in the cases of the influence of immigration and cultural factors, some impact of these milestones in Hungarian political life were visible, as for example, the effect of immigration sentiments went up after the refugee crisis and the xenophobic campaigns of Fidesz.
Such stability matches the findings of several authors who claimed, that despite the public perception of it, public stable phenomenon, with a limited number of extreme changes in it (Page & Shapiro, 1992; Druckman et al., 2012; Bartels & Bermeo, 2015).
Despite the fact, that my findings seem to be well supported and established, some remarks have to be made. My choice of variables might have impacted the results I found, or I might have excluded important factors yet to be found, that could have had more impact on Hungarian Euroscepticism. Lack of data in the case of my second dependent variable also might have had a negative impact on the accuracy of my analysis. In terms of future research, my work shed light on important gaps in the literature regarding Hungarian individual Euroscepticism. Researches like this one, with longer timeframes are of importance to continue, as I think without these, a compact and accurate picture on what drives anti-EU feelings in this country can only be drawn with huge difficulty. In addition, similar research regarding these sources of Euroscepticism before the early 2000s would be equally important. As I mentioned my research might have excluded country-specific sources of negative feelings towards the European Union, that could also be stronger predictors. Therefore, research on finding these Hungary specific factors would also be substantially important. For example, further researches should also focus
24
on variables regarding the regime change of 1989, as those who were raised before that can have completely different views. Researches that examine different subgroups of the Hungarian society would also contribute to a deeper understanding of how public Euroscepticism work in Hungary.
25 Appendix:
Correlation tables:
1.) Pearson correlation coefficients between Trust in the EP and the independent-, and control variables:
Additional Table 1. Source: European Social Survey (2002-2016)
Trust in the EP 2002 2004 2006 2008 2010 2012 2014 2016
Immigration 0,218 0,227 0,2 0,231 0,253 0,206 0,145 0,283 Stance on Gay Rights -0,123 -0,076 -0,032 -0,147 -0,068 -0,065 -0,144 -0,105 Political Trust Index 0,59 0,557 0,582 0,608 0,556 0,481 0,486 Economic Views 0,009 0,08 0,003 0,072 0,027 0,183 0,092 0,059 Gender 0,013 0,003 -0,025 0,056 0,05 0,048 0,054 0,032 Age -0,062 -0,037 0,022 -0,081 0,015 -0,037 -0,006 -0,03 Lower than Secondary
Education
-0,074 -0,03 -0,035 0,028 -0,017 -0,029 0,031 Secondary Education -0,052 -0,093 -0,085 0,001 -0,005 -0,009 -0,047 -0,039 Upper Secondary Education 0,016 -0,012 0,015 -0,067 -0,012 -0,056 -0,028 -0,043 Postsecondary Education 0,056 0,013 -0,029 0,009 0,044 -0,02 0,044 Tertiary Education 0,069 0,101 0,101 0,091 0,042 0,064 0,102 0,054 Household Income 0,101 0,043 0,105 0,224 0,015 Belonging to a Religion -0,039 -0,031 -0,038 0,025 -0,048 -0,03 -0,067 0,073 Level of Religiousness 0,045 0,079 0,023 0,055 0,081 0,184 0,05 0,011
2.) Pearson correlation coefficients between the European Unification variable and the independent-, and control variables:
Additional Table 2. Source: European Social Survey (2004-2016)
European Unification 2002 2004 2006 2008 2010 2012 2014 2016
Immigration 0,206 0,202 0,144 0,288 0,192 0,391 Stance on Gay Rights -0,132 -0,081 0,177 -0,129 -0,177 -0,32 Political Trust Index 0,151 0,201 0,127 0,15 0,018 -0,062 Economic Views 0,024 0,082 0,007 0,141 0,09 0,092 Gender -0,013 -0,051 0,031 0 -0,003 -0,009 Age -0,056 -0,019 -0,022 -0,053 0,1 -0,024 Lower than Secondary
Education
-0,05 -0,063 0,035 0,012 -0,035 0,087 Secondary Education -0,04 -0,038 -0,015 0,001 -0,019 0,029 Upper Secondary Education 0,013 0,007 -0,012 -0,099 0,003 -0,131 Postsecondary Education -0,021 0,027 -0,012 0,025 0,044 0,048 Tertiary Education 0,056 0,044 0,024 0,096 -0,002 0,074 Household Income -0,018 0,082 0,08 -0,05 Belonging to a Religion 0,026 0,035 0,091 0,085 0,004 -0,022 Level of Religiousness 0,003 -0,004 -0,025 -0,009 0,006 0,115
26 Graphs of Correlation Coefficients:
Additional Graph 1. Source: European Social Survey (2002-2016)
Additional Graph 2. Source: European Social Survey (2004-2016)
Intercorrelation table for Political Trust Index (Trust in Country’s Parliament, Trust in Politicians, Trust in Political Parties):
Source: European Social Survey (2002-2016)
First Round (2002) Second Round (2004)
Trust variables Parliament Politicians Parties Trust variables Parliament Politicians Parties
Trust in Parliament 1 0,636 Trust in Parliament 1 0,681 0,635
Trust in Politicians 0,636 1 Trust in Politicians 0,681 1 0,835
Trust in Political parties 1 Trust in Political parties 0,635 0,835 1
-0.5 0 0.5 1
Pearson Correlation coefficients (Trust in
the EP)
Immigration makes the country worse/better Gays and lesbians are free to live as they wish Political Trust Index
Government should reduce income inequalities
-0.4 -0.2 0 0.2 0.4 0.6
Pearson Correlation coefficients
(European Unification)
Immigration makes the country worse/better Gays and lesbians are free to live as they wish Political Trust Index
