Jump to ContentJump to Main Navigation
Show Summary Details
More options …

The B.E. Journal of Economic Analysis & Policy

Editor-in-Chief: Jürges, Hendrik / Ludwig, Sandra

Ed. by Auriol , Emmanuelle / Brunner, Johann / Fleck, Robert / Mendola, Mariapia / Requate, Till / Schirle, Tammy / de Vries, Frans / Zulehner, Christine

4 Issues per year


IMPACT FACTOR 2016: 0.252
5-year IMPACT FACTOR: 0.755

CiteScore 2016: 0.48

SCImago Journal Rank (SJR) 2016: 0.330
Source Normalized Impact per Paper (SNIP) 2016: 0.526

Online
ISSN
1935-1682
See all formats and pricing
More options …
Volume 16, Issue 2 (Apr 2016)

Issues

Volume 6 (2006)

Volume 4 (2004)

Volume 2 (2002)

Volume 1 (2001)

Euroskepticism, Income Inequality and Financial Expectations

Jo Ritzen
  • Institute for the Study of Labor (IZA), Bonn, Germany and Maastricht University, Maastricht, The Netherlands
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Caroline Wehner
  • Institute for the Study of Labor (IZA), Bonn, Germany and Maastricht University, Maastricht, The Netherlands
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Klaus F. Zimmermann
Published Online: 2015-12-05 | DOI: https://doi.org/10.1515/bejeap-2015-0052

Abstract

Before the Great Recession, the rising income inequality within the “old” European Union has been suggested as an important driver of the increase in Euroskepticism. We revisit this finding for the 27 EU member states from 2006 to 2011, introducing individual negative financial expectations as a further driving factor. We also distinguish between Western and Eastern European countries. In the period of Eastern EU enlargement after 2005, Euroskepticism increased by one third while income inequality on average remained stable. Negative financial expectations are positively related to Euroskepticism in the West and non-significantly negatively related in the East. This suggests that Westerners interpret European integration as a threat, while Easterners view it as a chance. In addition, income inequality lost its role in “old” Europe. An increase of one Gini point decreases the probability of Euroskepticism by half a percentage point in the West, while it has no impact in the East.

Keywords: Euroskepticism; income inequality; expectations; economic growth; unemployment

JEL: D31; J31; O43; O52; P48; Z18

1 Introduction

Euroskepticism is the European catchword for skepticism about the European Union. 1 Our attempt at understanding Euroskepticism is driven by its policy relevance. The policy relevance comes from the potential impact of Euroskepticism on the member states’ willingness to agree on further cooperative steps or enlargement. Euroskepticism also translates into the growth of anti-European parties and in a shift of traditionally pro-European parties toward a less pro-European cooperative point of view, thus undermining the foundations for strengthened cooperation between European countries.

We aim to contribute to the literature by shedding light on Euroskepticism formation in times of the recent financial and economic crisis and by answering the following questions: How stable are the trends from the past in explaining Euroskepticism? What is the relation between negative financial expectations and Euroskepticism? And are there differences between Western and post-communist EU member states? Our measure of Euroskepticism is based on the positive, neutral or negative answer to the Eurobarometer (EB) question, “Generally speaking, do you think that (your country)’s membership of the European Union is...?” We distinguish between Western Europe and the post-communist Eastern EU member states, as fundamental variation in the political and economic pathways in Western and former communist EU member states make it likely that the two regions have different decision-making processes about Euroskepticism.

We analyze Euroskepticism formation by following the utilitarian explanatory approach. In this approach individual attitudes are driven by socio-economic background variables such as gender, age, education, profession, and the degree of urbanization, as well as macro-economic variables pertaining to the country, namely GDP per capita, unemployment, inflation, income distribution and transfers from or to the EU. Before the Great Recession that started in 2007, the rising income inequality within the “old” European Union has been suggested as an important driver of the increase in Euroskepticism (Kuhn et al. 2014). This is in line with the concerns about the rise in inequality expressed in the broad public debate about the recent book by Piketty (2014). However, re-examining the data for the comprehensive period of 2006–2011 for the EU-27 countries, we cannot confirm that higher inequality drives Euroskepticism. To the contrary, we find the relationship to be negative in the West, while for the East there is no effect at all.

We further add to the literature by analyzing differences in the relation between utilitarian considerations and Euroskeptic attitudes in post-communist and Western EU countries within the period 2006–2011 by using the transmission mechanism of negative financial expectations. We thereby assume that those who are pessimistic about their financial future are particularly hit hard by the recent economic and financial crisis and that this translates into Euroskepticism differently depending on the region where people live. There is increasing awareness that economic behavior in the Great Recession cannot be fully explained with traditional models. Summers (2013) uses the variable “financial panic,” while Mau, Mewes, and Schöneck (2012) discuss the concept of socio-economic insecurity. Here we use the variable “financial expectations” as the individual evaluation of the socio-economic environment, testing whether this serves as a transmission mechanism for socio-economic variables toward Euroskepticism. We find that negative financial expectations significantly increase Euroskepticism in Western EU member states, but not in the newly accessed Eastern EU states.

The paper is structured as follows: In Section 2, we discuss the related literature, general trends, and our theoretical considerations. In Section 3, we present and discuss the model. The data are introduced in Section 4 and the model is empirically investigated in Section 5. Section 6 concludes and discusses implications for EU integration policies.

2 Explaining Euroskepticism

In a first step, we present the main explanatory approaches for Euroskepticism 2 and discuss differences between post-communist and Western EU member states with regard to public support for the EU. We also examine the implications of the recent economic and financial crisis as well as redistributive concerns for Euroskeptic attitudes. In a second step, we provide the motivation of our empirical strategy by discussing general trends and theoretical considerations.

2.1 Previous Findings

Euroskepticism is typically defined as “negative attitudes towards the EU and/or European integration” (Serricchio, Tsakatika, and Quaglia 2013, 52). Literature on Euroskepticism distinguishes between three main explanatory approaches for Euroskepticism: National identity, national institutional performance, and utilitarian theory (Loveless and Rohrschneider 2011). 3 The national identity approach explains resistance against European integration via feelings of an exclusive national identity and the fear of losing one’s own cultural identity (Diez Medrano 2010; Hooghe and Marks 2005, 2007; Lubbers and Jaspers 2011; McLaren 2002, 2007; de Vreese and Boomgaarden 2005). However, there is also evidence that strong national identity can exist in accordance with public support for European integration, which is known as inclusive national identity (Bruter 2005).

The national institutional performance approach explains attitudes toward the EU through the individual’s trust level toward national institutions. Thus, trust in national institutions is seen as a proxy for trust in European institutions, because citizens are much more informed about national politics than about the EU (Anderson 1998). In contrast, Sanchez-Cuenca (2000) argues that low trust in national political institutions can be substituted by high trust in EU institutions.

Finally, according to utilitarian theory, individual support for EU integration is positively associated with individual economic benefits that could be gained from EU market liberalization. Sociotropic utilitarianism regards national economic performance measured in GDP growth, inflation, unemployment, intra-EU trade or country net benefits from EU membership as decisive in shaping attitudes toward the EU (Anderson and Reichert 1995; Anderson and Kaltenthaler 1996; Eichenberg and Dalton 1993). Egocentric utilitarianism considers the individual socio-economic position measured by age, education level, and occupation as pivotal for the decision, because these characteristics are considered to be essential for being an economic winner or loser from EU integration (Gabel 1998a, 1998b; Gabel and Palmer 1995; Gabel and Whitten 1997). However, there is evidence that indeed the self-characterization as an economic winner or loser determines Euroskeptic attitudes, but that this self-characterization only partly overlaps with the individual socio-economic position. Thus, in this respect, individual attitudes toward Europe are more based on an assessment about how the EU affects someone personally (Mau 2005).

Before the Maastricht Treaty, the EU was primarily considered a project of economic integration. In this context, utilitarian theory was regarded as the dominant explanatory approach. With the Treaty, the popularity of the EU was used to enforce economic discipline among member states (Rotte and Zimmermann 1998). After establishing it, the EU expanded its competences into non-economic policy areas and then saw public support for EU integration decline, despite favorable economic conditions (Franklin and Wlezien 1997). At this point, the importance of national identity and national institutional performance explanatory approaches increased in Western European countries (Loveless and Rohrschneider 2011).

Looking at post-communist EU member states, Loveless and Rohrschneider (2011) state that the strongest determinants for positive attitudes toward EU integration in post-communist countries before EU accession are attitudes toward democracy and capitalism as well as the belief that the EU guarantees reforms (Kucia 1999; Cichowski 2000; Rohrschneider and Whitefield 2004). After EU accession, the importance of economic considerations increased. In accordance, Herzog and Tucker (2010) find that economic winners of the transition process are less Euroskeptic than losers. Comparing attitudes toward EU integration between East and West, de Vries (2013) argues that individuals in Western EU member states are more ambivalent toward EU integration than Eastern citizens. She explains that this difference stems from citizens in Western EU countries being more experienced with regard to positive and negative consequences of EU integration over the years.

After the onset of the financial crisis in 2007, Euroskeptic attitudes have increased considerably among the EU-27 countries. Following proxy mechanism theory by Anderson (1998) and mainly denying the importance of utilitarian considerations, Armingeon and Ceka (2014) find that EU attitudes are mainly derived from evaluating the national government. They conclude that if national governments are successful in solving economic problems based on the crisis, the support for the EU will increase again. In addition to this, Serricchio, Tsakatika, and Quaglia (2013) consider exclusive national identity as decisive for Euroskepticism. Levy and Phan (2014) take a more integrative point of view by stating that the sociotropic assessment of the national economic situation drives EU attitudes, particularly among those with an exclusive national identity. They conclude that if the economic crisis produces a resurgence of nationalism, the national economic situation becomes even more important to assure the project of European integration. Finally, Braun and Tausendpfund (2014) show that contrary to the predominant opinion, utilitarian considerations again play an important role in explaining attitudes toward the EU.

Besides the explanatory approaches discussed above, Eichenberg and Dalton (2007) argue that redistributive concerns are crucial for attitudes toward the EU. There is evidence that European economic and political integration is one important driver of an increase in income inequality (Beckfield 2006, 2009) and that the increase in income inequality is negatively associated with public support for EU integration (Burgoon 2013; Kuhn et al. 2014). Recent findings furthermore indicate that individuals with lower levels of education are particularly sensitive to income inequality with regard to Euroskeptic attitudes (Kuhn et al. 2014) and that the positive impact of low education on Euroskepticism has even increased in the last decades (Hakhverdian et al. 2013).

2.2 General Trends and Theoretical Considerations

Figure 1 shows that Euroskepticism has increased considerably from 2006 to 2011. Based on the reviewed literature, we conclude that utilitarian evaluations are important for individual attitudes toward the EU, particularly during the European economic and financial crisis (Braun and Tausendpfund 2014). However, we hypothesize that individual financial evaluations affect attitudes toward Europe in post-communist and Western EU countries differently. We aim to contribute to the literature by further shedding light on this issue by answering the following questions: How stable are the trends from the past in explaining Euroskepticism? What is the relation between negative financial expectations and Euroskepticism? And are there differences between Western and post-communist EU member states? To our knowledge there is no article that analyzes differences in the relation between utilitarian considerations and Euroskeptic attitudes in post-communist and Western EU countries during the crisis period by using the transmission mechanism of negative financial expectations. We aim to fill this research gap.

First we look at the question of general trends with particular emphasis on income inequality. Motivated by the finding that European integration divides European citizens into economic winners and losers (Beckfield 2006), one prominent explanatory approach for an increase in Euroskeptic attitudes is the increase in income inequality (Atkinson 2013; Burgoon 2013; Kuhn et al. 2014; Ritzen and Zimmermann 2014). However, for the period of 2006–2011 (i.e. the period covering the economic crisis), we do not observe a strong increase in income inequality in Western EU countries, while post-communist countries even experienced a decrease in income inequality (see Figure 2). This development is presumably related to the fact that within the crisis the employment rates declined across the income distribution. This means that former middle and high wage earners who were previously winners of European integration might have been negatively affected by the EU financial and economic crisis. Thus it is not so surprising if income inequality is not related to, or even negatively related to, Euroskepticism.

Development of income inequality (measured by Gini) between 2006 and 2011.Source: Eurostat Database 2013a. Own calculation of unweighted percentages.
Figure 2:

Development of income inequality (measured by Gini) between 2006 and 2011.

With regard to financial expectations, we assume that those who have negative financial expectations are hit particularly hard by the crisis. This group not only comprises citizens who had already been economically disadvantaged before the crisis, but presumably also includes former economic winners of European integration. To generalize from this, below we introduce and explain the variable negative financial expectations as a transmission mechanism in our analysis. Economic sentiments are often seen as a transmission mechanism between real-world variables and economic decisions (e.g. Beckmann, Belke, and Kühl 2011). Introducing negative financial expectations is a way to incorporate the sentiment of economic uncertainty into the Euroskepticism explanation. Thereby, financial expectations are hypothetically driven by economic circumstances, which are partly affected by the crisis and translate into Euroskepticism. Figure 3 descriptively supports our approach by showing that negative financial expectations strongly increase after the onset of the financial and economic crisis in 2007.

The Great Recession that started in 2007 implies a huge degree of economic uncertainty for EU citizens. We assume that in this context individual financial future expectations, which are based on individual and national economic circumstances, are important for citizens’ attitudes toward the EU. This is because EU politics, in addition to national politics, have played an important role in managing the financial and economic crisis (Serricchio, Tsakatika, and Quaglia 2013).

We hypothesize that in Western EU member states citizens with negative financial expectations are more likely to report Euroskeptic attitudes than those who believe that their economic situation will not change and those who have positive financial future expectations. People with negative financial expectations are mostly from countries that have been hit hardest by the crisis such as Greece and Portugal. Under the pressure of EU policies, these countries have had to impose stability and reform measures that have been highly unpopular. Such measures, sometimes summarized critically as austerity can be perceived by some, at least for the short term, as a further threat to the financial future, particularly for those already negatively affected by the crisis. This may result in an increase in Euroskeptic attitudes (see Braun and Tausendpfund 2014, but also Mau 2005). However, we do not deny that this relation is possibly also affected by the lack of trust in national and EU institutions (Armingeon and Ceka 2014; Serricchio, Tsakatika, and Quaglia 2013).

In contrast, we expect financially pessimistic individuals in post-communist countries to be much more reluctant with regard to Euroskepticism than citizens from Western EU member states. We neither want to deny that people in Eastern EU countries with lower education or lower occupation statuses are more likely to have Euroskeptic attitudes than for instance highly educated managers (Herzog and Tucker 2010), nor do we claim that citizens from post-communist countries have not also had to face hard austerity policies. However, we believe that in post-communist countries, Europe is still linked to popular political and economic reforms (Cichowski 2000; Herzog and Tucker 2010; Kucia 1999; Rohrschneider and Whitefield 2004) as well as to the experience of economic convergence and growth based on liberalized markets and EU transfers (Gill and Raiser 2012). Furthermore, Eastern EU member state citizens have less experience with the disadvantages of EU policies (de Vries 2013). We therefore expect that in post-communist countries negative financial expectations are much less related to Euroskeptism than in the Western EU, and that the EU is still seen as a solution for rather than a source of economic problems. Figures 1 and 3 support our argumentation since they clearly show that although post-communist countries have a higher share of citizens with negative financial expectations than those in the West, they are less likely to express Euroskeptic attitudes. We empirically test our hypothesis throughout the next sections.

3 Model

Our empirical investigation seeks to identify the impact of individual negative financial expectations and a vector of micro- and macro-economic variables on Euroskepticism. The variable “negative financial expectations” is a binary variable for either having or not having negative financial expectations. Euroskepticism is also a binary variable for being or not being Euroskeptic. It is possible that there are unobserved variables that make people both more likely to have negative financial expectations and to be Euroskeptic. To account for the recursive structure of our approach and for possible unobserved jointly exogenous variables, we estimate a recursive bivariate probit model (RBP).

Based on the theoretical considerations, we hypothesize that negative financial expectations have a positive impact on Euroskepticism in Western EU countries and no (or a negative) effect in Eastern EU member states. Furthermore, we assume that these expectations serve as the transmission mechanism for the mood created by economic circumstances. The transmission process means that both Euroskepticism and individual negative financial expectations are jointly determined by variables within a recursive structure so that the error terms of eq. [3.1] and [3.2] might be correlated: Euroskepticismi,j=α0+α1negativefinancialexpectationsi,j+α2nationalmacrovariablesj+α3EUbudgettransfersj+α4socioeconomicbackgroundi,j+ε1[3.1] Negativefinancialexpectations(i,j)=β0+β1nationalmacrovariablesj+β2EUbudgettransfersj+β3socioeconomicbackgroundi,j+ε2[3.2]

National variables are income inequality (measured by the Gini coefficient), unemployment, GDP per capita and inflation (HICP). The EU budget transfers are net calculations and in relation to the country’s gross national income (% GNI). We hypothesize that increases in income inequality lead to an increase in Euroskepticism. Unemployment and inflation are indicators of income uncertainty and are hypothesized to lead to increased Euroskepticism. An increase in GDP per capita is an indicator of income gains and is expected to decrease Euroskepticism. Higher net transfers received from the EU are expected to lead to more support for the EU and to less Euroskepticism.

There are likely to be future variables explaining negative financial expectations in eq. [3.2], but those are either not available or if available they would be highly endogenous. We prefer to concentrate on predetermined variables explaining the future, and control for the unobserved heterogeneity by modeling the correlation across both equations of the system. In that sense, eq. [3.2] “instruments” the endogenous variable negative financial expectations in eq. [3.1], and allows us to estimate the effects of the predetermined variables on both endogenous variables separately.

Modeling the feedbacks from the effects of Euroskepticism to the sub-system of respondents’ attitudes admittedly would go beyond the purpose of this paper. However, there is the potential that the general rise in Euroskepticism across Europe can affect individual attitudes and reach variables such as negative financial expectations and even have an effect on real world factors like unemployment and growth.

Negative financial expectations will be determined by information about the future, which variables included in the past do not cover. However, this lack of coverage is an advantage since this reduces the degree of endogeneity the “instrumented” variable negative financial expectations exhibits. This also holds for a potential effect of negative financial expectations on future macro-variables, which eq. [3.2] does not cover. We also consider residual correlation to take into account unobserved heterogeneity. Euroskepticism may well drive “financial expectations” because it may decrease the possibility of debt mutualization in the Euro area. However, individuals know that those effects are pretty slow due to the sluggish political process. Hence, effects might only be marginal in the short term.

For the period before the economic and financial crisis and regressing the level of Euroskepticism on the change of income inequality in the EU-12 countries, Kuhn et al. (2014) found a positive relationship. This implies that a positive change of income inequality in one period resulted in a higher level of Euroskepticism. Differently, we model the levels of Euroskepticism and negative financial expectations as a function of inequality measured by the Gini coefficient. Both different specifications have well-known different empirical implications. A one period increase in the Gini has only a temporary effect on Euroskepticism in the Kuhn et al. (2014) specification, if the increase does not repeat. However, in our specification, an increase in the level of inequality induces a permanent rise in Euroskepticism even if the Gini remains fixed over the next periods. We find that our level representation is more appropriate for our data, but further discuss and examine the empirical differences in Section 5.2.

Our structural estimation approach directly models the endogeneity generated by potentially correlated error terms in the recursive equation system [3.1] and [3.2]. The model also allows us to decompose the impact of the exogenous regressors on Euroskepticism in a direct effect through eq. [3.1] and an indirect effect through eq. [3.2]. This also enables us to examine the estimates’ consistency from the recursive model with those from a final form probit model directly estimating the total effects, replacing negative financial expectations in eq. [3.1] by eq. [3.2] and collecting the regressor terms.

To calculate unbiased joint estimates of the two processes, we estimate a RBP model that simultaneously estimates the probability of being Euroskeptic conditional on the probability of having negative financial expectations. This bivariate probit model (Maddala 1983, 122–3) is formulated as follows: y1i=xiτ1+y2iπ+μ1iy1i=1,ify1i>0,0otherwise,[3.3] y2i=xiτ2+μ2iy2i=1,ify2i>0,0otherwise,[3.4]where y*1i is the latent variable associated to the binary dependent variable Euroskepticism of eq. [3.1]; y*2i is the latent variable associated to the binary dependent variable negative financial expectations of eq. [3.2], which is included in eq. [3.1] as an binary endogenous variable; xi includes the two regression equations’ exogenous regressor vectors; and µ1i and µ2i are the error terms. We assume that the error terms µ1i and µ2i are standard normally distributed (N (µ, σ2)=N (0, 1)) and that covariance of the error terms equals Cov(µ1i, µ2i | x1i, x2i)=ϱ. If the error terms of the two equations are uncorrelated, i.e., ϱ=0, then both equations can be estimated separately. But, if the error terms are correlated and ϱ ≠ 0, separately estimated results would be biased. Identification in the RBP model as formulated above can rely alone on the functional form based on non-linearity (Green 2003; Wilde 2000).

The model is separately applied to the EU-27 as a whole, to Western (17) and to Eastern EU (10) countries in order to trace whether the political and economic pathways in Western and former socialist EU member states imply different decision making processes regarding Euroskepticism. Western EU countries include Austria, Belgium, Cyprus, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Malta, The Netherlands, Portugal, Spain, Sweden, and the United Kingdom. The former socialist EU member states are Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, and Slovenia. The Eastern EU countries are all recent members to the EU, while most of the Western countries were already part of the EU in 2004 (except for Malta and Cyprus, which joined in 2004). Eight Eastern European countries (the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia) accessed the EU in 2004, while Bulgaria and Romania joined in 2007.

4 Data

The analysis is based on a pooled cross-sectional dataset with detailed micro- and macro-data for the 27 EU member states for the period 2006–2011. Individual data include attitudes toward one’s own country’s EU membership, individual financial expectations, as well as demographic and socio-economic information provided by the cross-sectional Standard Eurobarometer (EB) survey. Eurostat provides the country-specific macro-indicators. These consist of the Gini coefficient as a measure for income inequality, annual unemployment rate averages, gross domestic product (GDP) per capita, the harmonized consumer price index (HICP) measuring inflation, and EU net transfers in relation to the gross national income (% GNI).

Our key variables are Euroskepticism and Negative Financial Expectations. Euroskepticism is measured by the EB question, “Generally speaking, do you think that your country’s membership of the European Union is…?” with answer categories: (1) a good thing, (2) a bad thing, (3) neither good nor bad, or (4) don’t know (DK). We create the binary variable Euroskepticism with response categories (1) a bad thing and (0) a good thing or neither good nor bad. Category (4) responses are treated as missing values. Financial expectations are inquired by the EB question, “What are your expectations for the next twelve months: will the next twelve months be better, worse or the same, when it comes to the financial situation in your household?” with the response options: (1) better, (2) worse, (3) same, or (4) don’t know (DK). We recode the binary variable Negative Financial Expectation with categories (1) worse and (2) better or same. “Don’t know” answers are treated as missing values.

After merging the eight relevant EB waves (65.2, 67.2, 70.1, 71.1, 71.3, 72.4, 73.4, 75.3), the dataset originally consisted of 213,633 observations. Focusing on the economically active population reduces the number of observations to 147,129. We dropped citizens who retired early and those older than 64 because they may have had a different process of Euroskepticism formation. Missing values for our key variables Euroskepticism and Negative Financial Expectations account for a loss of 7,878 observations. Missing values in any other of the micro-variables lead to a loss of a further 1,913 observations. Therefore, our analysis is based on 137,338 observations that consist of 85,881 observations for Western European countries and 51,457 for former socialist EU member states.

In Table 1, we present our sample’s descriptive statistics for the EU-27. This information is separated for the two regions distinguished in the Appendix, Tables 7 and 8. The average age of individuals in the Western countries sample is slightly older, they have more years of education, there are more house persons, less unemployed, and less live in large towns. The occupation variable “house persons” describes individuals who are responsible for the household and domestic tasks and who are inactive in the labor market. The macro-variables show, in particular, the sizeable difference in GDP per capita (higher in the West) and in transfers from the EU (higher in the East).

Table 1:

Descriptive statistics, EU-27, 2006–2011 (N=137,338).

4.1 Euroskepticism

Euroskepticism in the EU increased from 12% in 2006 to 17% in 2011. In 2011, the Euroskeptics still formed less than one fifth of the population. At the same time, the group has increased by almost one third. The EU-wide figures mask substantial differences between countries. In 2011, the countries with the most Euroskepticism were Greece (32%), Portugal (29%), Cyprus (27%) and the United Kingdom (26%); those with the least included Bulgaria, Estonia, Poland, Slovakia, and Belgium (below 11%). Between 2006 and 2011, only Finland, Sweden and Estonia saw a slight decrease in Euroskepticism (maximum decline of 3 percentage points). In contrast, between 2006 and 2011, many countries showed a sharp increase (presented in percentage points): notably, Greece (20), Slovenia (16), Portugal (14), Spain (11), Hungary (11), Cyprus (9), Latvia (7), Italy, Lithuania, Luxembourg and the United Kingdom (6). All of the countries that had applied for EU Emergency Support are among those showing a sharp increase. More moderate increases are found in Denmark, France, Germany, Ireland, Malta, The Netherlands, Poland, Romania, and Slovakia (maximum increase of 4 percentage points).

4.2 Income Inequality

The increase in the Gini coefficient witnessed in the OECD (2008, 2011) for the period 1975–2005 did not take place in most EU countries in the period 2006–2011. The Gini (multiplied by 100) increased in the following countries: Austria (by 1 Gini point), Bulgaria (3.8), Cyprus (0.4), Germany (2.2), France (3.5), Malta (0.4), Romania (0.2), Slovenia (0.1), Spain (2.6), Sweden (0.4) and the United Kingdom (0.5). The highest increase was in Denmark (4.1). The Gini decreased in: Belgium (by 1.5 Gini points), the Czech Republic (0.1), Estonia (1.2), Finland (0.1), Greece (0.8), Ireland (2.1), Italy (0.2), Latvia (3.8), Lithuania (2), Luxembourg (0.6), the Netherlands (0.6), Poland (2.2), Portugal (3.5) and Slovakia (2.4). The highest decrease, 6.5 Gini points, was observed in Hungary.

4.3 Financial Expectations

The share of people in the EU-27 who believe that their personal financial situation will worsen increased from 17% in 2006 to 27% in 2008 and then decreased again to 19% in 2011. In 2011, negative financial expectations were highest in Greece (54%), Portugal (42%), Hungary (32%) and Romania (31%). Countries with the lowest share of pessimistic citizens in 2011 comprise the Scandinavian countries and Luxembourg, with percentages below 9%. On average, around 20% of Europeans in the sample express gloomy financial prospects. The share of citizens with pessimistic financial expectations in the pooled dataset is slightly higher in Eastern Europe (25%) than in Western EU member states (19%) over the six-year period between 2006 and 2011.

4.4 Increasing Unemployment

Many (more than expected) EU-27 countries managed a decrease in unemployment between 2006 and 2011: Austria (from 4.8% in 2006 to 4.2% in 2011), Belgium (8.3% to 7.2%), the Czech Republic (7.1% to 6.7%), Germany (10.3% to 5.9%), Malta (6.9% to 6.5%) and Poland (13.9% to 9.7%). Yet the crisis hit employment hard in many other countries. The highest unemployment rate and also the highest increase of unemployment was found in Spain (8.5% in 2006 to 21.7% in 2011), followed by Greece (8.9% to 17.7%), Latvia (6.8% to 16.2%) and Lithuania (5.2% to 15.4%).

4.5 GDP per Capita

GDP per capita rose in almost all of the EU-27 countries in the period 2006–2011, with a relatively fast and steady growth in the Eastern European countries. There was a decline only in Ireland and the United Kingdom, while the Southern and North-Western European countries remained more or less at a standstill. In 2011, the highest GDP per capita was observed in Luxembourg (€80,300), followed by Denmark (€43,200), while the lowest was in Bulgaria (€5,200) and Romania (€6,100).

5 Results

5.1 Euroskepticism Explained by Financial Expectations

In Table 2, we present the marginal effects of the RBP regressions for the overall sample and separately for Western and former communist Eastern EU member states. We control for country and year fixed effects to account for country and time-specific factors. We also calculate robust and country clustered standard errors to account for heteroskedasticity and correlated error terms within countries. The country and time dummies remove a large amount of the variance, allowing us to concentrate on the factors that are related to the effects we want to study. We also investigate the robustness of our findings using alternative estimation techniques in Section 5.2.

We hypothesize that financial expectations act as a transmitter of socio-economic circumstances toward Euroskepticism. The results support our approach by showing that negative financial expectations have a highly significant positive effect on Euroskepticism in the Western EU countries, but a non-significant negative effect on Euroskepticism in the post-communist countries. This result suggests that there are different mechanisms leading to Euroskepticism in the two regions. Euroskepticism has always been higher in the West than in the East (see Figure 1), and the reverse has been true for negative financial expectations (see Figure 3). But while citizens in the West were also blaming Brussels for the threat of an increased financial burden, members in the East were not expecting additional burden from the EU, but perhaps support due to the economic situation in those countries.

The last row of Table 2 (Wald test) shows considerable variation in the correlations of the disturbances of the eqs [3.1] and [3.2] between the two regions. For each region the first number shows the estimated correlation of the disturbances of eqs [3.1] and [3.2] while the second number presents the significance level of the likelihood-ratio test (which tests the null-hypotheses of uncorrelated error terms of ϱ=0). In Western EU countries we find a non-significant negative correlation of –0.1627 (p<0.2136). Therefore, Euroskepticism and negative financial expectations may not be jointly determined in Western EU countries and two separate probit regression models should give similar results. However, in Eastern EU countries the disturbances of eqs [3.1] and [3.2] are borderline significantly positively correlated (p<0.1372). The fact that the estimated correlation of the disturbances is with 0.6615 substantially different from zero suggests that it is important to control for endogeneity in Eastern EU countries. The estimation of a RBP model is therefore appropriate to get unbiased estimates. The result further supports our expectation that Euroskepticism in the two regions follow different decision-making processes.

First, we inspect the macro-explanatory variables. Contrary to our expectations and to all previous studies, income inequality measured by the Gini coefficient has a statistically negative impact on Euroskepticism and on negative financial expectations in Western EU countries. An increase in income inequality by one Gini point decreases the probability of being Euroskeptic by 0.5 percentage points (pp) and of having negative financial expectations by 1.7 pp. This finding shows that the results obtained by Kuhn et al. (2014) for the EU-12 for the period 1976–2008 no longer hold for income inequality in Western Europe.

Income inequality has no significant effect in post-communist countries on Euroskepticism or negative financial expectations. The unemployment rate boosts negative financial expectations only in Western European countries, not in Eastern Europe. A one percentage point increase in unemployment increases the probability of having negative financial expectations by 1.2 pp in Western Europe. Inflation slightly increases negative financial expectations in both Western and Eastern EU countries, but has no significant association with Euroskepticism in either region. GDP per capita is an important determinant of Euroskepticism and negative financial expectations in Eastern Europe, but has no effect in Western EU countries. An increase in GDP per capita of 1% decreases the probability of being Euroskepticism by 0.25 pp and of having negative financial expectations by almost 0.40 pp in former socialist EU countries. Finally, EU net transfers are negatively associated with Euroskepticism in Eastern EU countries. Yet in Western EU countries, which on average are net payers of EU transfers, EU net transfers have no significant association to Euroskepticism.

Looking at the micro-explanatory variables we find that in particular low education and low occupation status are both positively associated with Euroskepticism and negative financial expectations both for Western as for Eastern EU countries. The results for the occupation dummies provide similar evidence. For example, in comparison to the reference group of managers, being unemployed increases the probability of being Euroskeptic by 9.5 pp in Western EU countries and by 8.8 pp in former socialist countries. Furthermore, being unemployed boosts the probability of having negative financial expectations by 8.2 pp in Western EU countries and by 13 pp in Eastern EU countries.

The data set has a hierarchical structure, where respondents are cross-nested in countries as well as years. We have just dealt with this by including dummies and by calculating country clustered and robust standard errors to permit heteroskedasticity and within-cluster error correlation. However, when the number of clusters is small (say 5–30) as in our samples, the standard asymptotic tests may over-reject resulting in too large standard errors for the variables at the cluster level. As Cameron, Gelbach, and Miller (2008) found in Monte Carlo estimations, bootstrap-based procedures can improve inference and lead to much lower rejection rates than standard methods. Table 9 provides a robustness check comparing for key variables the original standard error with those using the robust, cluster and bootstrap methods. As expected, clustered and robust standard errors are much larger than without, and the bootstrap standard errors are all close to the robust standard errors. Hence, the cluster-based evidence used in Table 2 is conservative, but using the less strict bootstrap standard errors would not lead to different conclusions in our case.

The hierarchical structure of our data could be also fruitfully analyzed by a multilevel analysis (MLM, see Gelman and Hill 2009, for instance). One reason why we do not present such analysis here is that we are not concerned with the additional information multilevel analysis may provide. Also clustering and multilevel analysis are basically equivalent methods as has been shown in Monte Carlo simulations by Harden (2009). However, as has been demonstrated recently in Monte Carlo studies by Stegmueller (2013) and Bryan and Jenkins (2015), estimates of parameters and standard errors for the cluster-level variables can be seriously biased when the cluster size is small (e.g. below 30), while the individual-level effects are reliable. Therefore, MLM is no panacea.

Table 2:

Recursive bivariate probit regressions, 2006–2011, Euroskepticism (EUS) and negative financial expectations (NFE).

5.2 Reduced Form Estimates

To test our results based on the RBP model, we estimate the reduced form for Euroskepticism and negative financial expectations based on separate probit models. The results of the probit regressions for the EU-27 as well as Western and Eastern EU member states are presented in Table 3. The probit regression results for the dependent variable Negative Financial Expectations vary only marginally from the RBP estimates. However, in comparison to the RBP results, the marginal effects of the explanatory variables on Euroskepticism alter their effect size in the probit regression, as expected, because they also take over the effect of the negative financial expectations, namely the factors driving them. Hence, these estimates are the explanatory variables’ total effects, while Table 2 had decomposed those into direct and indirect effects as discussed in Section 2.

We can regard the comparison between the first, third and fifth column of the probit analysis (Table 3) without negative financial expectations and the recursive RBP model’s results in Table 2 as confirmation of the transmission mechanism of financial expectations for the effect of financial expectations toward Euroskepticism. For instance, when we financial expectations, the size and significance of the Gini estimated effects on Euroskepticism for the total sample and the West in the regressions without financial expectations (Table 3) decreases in the RBP model (Table 2). For the post-communist EU, there is no effect in the reduced form (see Table 3, fifth column), and it remains insignificant in the system estimate (see Table 2, fifth column). Another example is unemployment, which is very significant in the Western EU in determining Euroskepticism in the reduced form (see Table 3, third column), but is statistically insignificant in the system estimate (see Table 2, third column). The fourth column in Table 2 reveals that the effect operates entirely through negative financial expectations, which unemployment significantly affects.

Again we first look at the macro-explanatory variables. Without taking the possible endogenous negative financial expectations into account, the statistically significant negative relation between the Gini coefficient and Euroskepticism is even increasing in Western EU countries. An increase in income inequality by one Gini point decreases the probability of being Euroskeptic by 0.8 pp (0.5 pp in RBP). In former socialist EU member states, we still find no significant effect. An increase of 1% in the unemployment rate now significantly boosts Euroskepticism by 0.5 pp in Western EU countries (insignificant in RBP) but has no effect in Eastern EU countries. Inflation still has no significant effect on Euroskepticism in either region. In Eastern Europe, GDP per capita is still significantly associated with Euroskepticism. However, the effect size is decreasing. An increase in GDP per capita of 1% decreases the probability of being Euroskeptic by 0.16 pp (0.25 pp in RBP). The same is found for EU net transfers, which are negatively associated with Euroskepticism in Eastern EU countries. A one percentage point increase of EU net transfers (% GNI) decreases the probability of being Euroskeptic by 1 pp (1.5 pp in RBP).

Looking at the micro-explanatory variables, we confirm the results that, in particular, low education and low occupation status are positively associated with Euroskepticism. However, we again find that neglecting negative financial expectations leads to a larger effect size in Western EU countries and a smaller effect size in Eastern EU countries. The results of the education dummies show that in comparison to the reference group that has obtained at least 20 years of education, having less than 16 years of education increases the probability of being Euroskeptic by 10.5 pp (9 pp in RBP) in Western EU countries and by 4.4 pp (6.7 pp in RBP) in Eastern EU countries. The results for the occupation dummies provide similar evidence. In comparison to the reference group of managers, being unemployed increases the probability of being Euroskeptic by 11.2 pp (9.5 pp in RBP) in Western EU countries and by 5.9 pp (8.8 pp in RBP) in former socialist countries.

Table 3:

Probit regressions, 2006–2011, Euroskepticism (EUS) and negative financial expectations (NFE).

5.3 Further Robustness Checks

To test the robustness of our results, we follow Kuhn et al. (2014) and re-estimate the RBP model by using changes in macro-economic variables. The procedure shows that Gini changes are not significantly related to Euroskepticism in Western EU countries, and slightly negatively in post-communist countries. Furthermore, removing the year dummies from the initial RBP model yields similar results as presented in Table 2, see Appendix Table 10. These findings confirm that there is no relationship between an increase in the Gini coefficient and Euroskepticism in our sample. The results deviating greatly from former findings may be because the Gini coefficient in most of the countries did not vary as much, or even decreased, in the observation period compared to before the Great Recession.

Second, we check robustness by including those who retired early and those who are older than 64 into the working sample. This changes the size of the working sample from about 140,000 to about 200,000 observations. However, Table 4 shows that our RBP estimation results are largely robust.

Table 4:

Robustness check: recursive bivariate probit, 2006–2011, including retired people, Euroskepticism (EUS) and negative financial expectations (NFE).

Finally, we estimate our RBP model for the time periods before and within the economic crisis separately. We consider the years 2006–2008 as the period before the economic crisis and the years 2009–2011 as the period within the economic crisis. Our results are presented in Tables 5 and 6. Looking at Western EU countries, the robustness check shows that having negative financial expectations increases the probability of being Euroskeptic by 15.6 pp before the economic crisis, while it is 19.5 pp during the crisis. In Eastern EU member states, the non-significant negative effect of having negative financial expectations on Euroskepticism in the pre-crisis period becomes significantly negative in the within-crisis period. Table 6 indicates that during the crisis, having negative financial expectations decreases the probability of being Euroskeptic by 14 pp in former socialist countries.

To test whether the difference between the coefficients for negative financial expectations before and within the crisis is statistically significant, we estimate the RBP model for the period 2006–2011 by including an interaction term between a crisis dummy and negative financial expectations. The coefficient for the interaction terms for both Western and former socialist countries reveals that the effect of negative financial expectations before the economic crisis (2006–2008) is significantly different from the effect of negative financial expectations within the economic crisis (2009–2011).

This result further supports our argument that financial expectations act as a transmitter of socio-economic circumstances toward Euroskepticism. The result suggests that there are different mechanisms leading to Euroskepticism in the two regions and that financially pessimistic people in Western Europe might interpret European integration as a threat to their financial situation, while Eastern European people might view it as a chance to improve their economic situation, particularly in times of economic decline. People in Northern European EU member states may be worried by high financial transfers to the European Union, while Southern EU countries might fear austerity measures. In contrast, Eastern EU countries may appreciate European integration due to positive net transfers or improved employment opportunities based on an integrated European labor market. However, the robustness check further shows that all other results stay mostly robust.

Table 5:

Robustness check: recursive bivariate probit, 2006–2008, Euroskepticism (EUS) and negative financial expectations (NFE).

Table 6:

Robustness check: recursive bivariate probit, 2009–2011, Euroskepticism (EUS) and negative financial expectations (NFE).

6 Discussions and Conclusions

Our study analyzes Euroskepticism formation within the period of 2006–2011 by emphasizing differences between Western and Eastern EU member states and by using the “negative financial expectations” transmission mechanism, thus covering those who have been hit particularly hard by the crisis. We provide evidence that having negative financial expectations determines Euroskeptic attitudes differently in Eastern and Western EU countries. In Western EU countries, we find a positive relation between negative financial expectations and Euroskepticism, while there is no significant relation in post-communist countries. This result suggests that in the period around the recent economic and financial crisis, Western EU citizens who are negatively affected by the crisis interpret European integration as a threat because they likely fear that austerity policies imposed by the EU further worsen their financial situation. In contrast, people from Eastern EU countries who are hit hard by the crisis are much more reluctant to adopt Euroskeptic attitudes because they still consider Europe as a source of solutions for economic problems. Thus, in post-communist countries, Europe is still much more connected to popular political and economic reforms as well as a source of economic convergence and growth based on liberalized markets and EU transfers. Eastern EU member state citizens have less experience with the disadvantages of EU policies, further contributing to higher levels of public support for the EU compared to those from Western EU countries.

With regard to sociotropic utilitarian evaluations, economic variables have explained Euroskepticism well, mainly for the “old EU” countries and in the pre-crisis period. For instance, Kuhn et al. (2014) study the EU-12 countries from 1976 to 2009 and show that Euroskepticism increased in a statistically significant manner, with more income inequality and higher unemployment but not with greater inflation. In this paper, we find that Euroskepticism formation has changed with regard to income inequality in the enlarged European Union, in both the old EU as well as in the new Eastern EU member states. Compared to Kuhn et al. (2014), we find a change in the sign of income inequality’s impact on Euroskepticism in Western Europe during the period of 2006–2011. In post-communist countries, income inequality has no significant relation to Euroskepticism. A possible explanation for this result is that inequality has on average only slightly increased in this period across the Western part of the EU but was, on average, falling in the Eastern part. The strong rise in Euroskepticism over this period does not match the more modest development of inequality. In spite of Piketty’s (2014) book, which has recently stirred a broad interest in inequality, our study does not show inequality as the direct driver of Euroskepticism. However, this is not Piketty’s line, nor does it rule out that inequality may moderate growth and hence contribute to weaker growth and rising unemployment, therefore affecting financial expectations and Euroskepticism.

We further find for Western EU countries that unemployment feeds negative financial expectations and affects Euroskepticism only indirectly. This result further supports our argumentation because it shows that the individual perspective of financial uncertainty arising out of increasing unemployment is more relevant for Euroskepticism than the unemployment rate per se. GDP per capita and transfers to other EU countries are of no concern. In post-communist EU countries, profiting from EU transfers and the rise in per capita GDP directly reduce Euroskepticism, reflecting positive perceptions of the EU’s ability to solve economic problems.

With regard to egocentric utilitarian considerations, the previous literature consistently shows that a disadvantaged socio-economic position is positively related to Euroskepticism. Looking at egocentric utilitarian variables, our results confirm these findings. More education is associated with a lower probability of having both negative financial expectations and Euroskeptic attitudes; contrarily, being unemployed is related to a higher probability of having negative financial expectations and of being Euroskeptic. However, the results show that unemployed people in the East are more likely to have negative financial expectations, but are less likely to be Euroskeptic than those from the West.

What are our predictions for the years after 2011? Member states are slowly emerging from the crisis with positive GDP growth rates in 2014 for all EU countries except Cyprus, Italy, and Finland. In Italy negative growth at least slowed down between 2012 and 2014 (Eurostat Database 2015). This development is likely to improve economic sentiments and moderate Euroskepticism, particularly in Southern EU member states (for initial descriptive evidence, see European Commission 2015a or Pew Research Center 2015). However, in line with our argumentation, as the economic situation in Cyprus still remains difficult, 42% of respondents have a negative image of the EU, which is the highest share of citizens with this attitude among all EU member states in 2015 (European Commission 2015a, 9). In Finland, a Euroskeptic political party came into office in 2015, perhaps advantaged by the persistent negative GDP growth.

Recently, EU membership popularity seems to have also increased as the conflict with Russia makes the EU more of a “safe haven,” particularly for some Eastern European countries. However, concerns about financial assistance to crisis countries may affect Eastern public support for the EU. For instance, facing growing anti-European sentiments, the Slovak government collapsed in 2011 because of its contributions to the European Financial Stability Facility (EFSF). In addition, the refugee crisis and its subsequent discussions regarding distributing asylum-seekers across EU countries, as well as the ongoing debates about possibly limiting intra-EU labor mobility, have also put substantial strain on positive European feelings. There is a need to further research each of these topics. Hence, our predictions are that economic evaluations will remain important in explaining high levels of Euroskepticism.

There are also policy implications emerging from our empirical results. Certainly we demonstrate that economic factors drive concerns about Europe. Hence, focusing on sustainable economic growth, reducing unemployment, and lowering financial insecurity will likely reduce those concerns. A rise in income inequality seems less likely to fuel fear about European membership. This aligns with a recent European Commission study that asks which topics should be emphasized in order to cope with global challenges and finds that “progress and innovation is gaining ground at the expense of social equality in many countries” (European Commission 2015b, 11). Finally, it is a problem that citizens have the feeling that their voice does not count in the EU political decision-making process, especially those in countries such as Cyprus or Greece where the economic situation is under pressure (European Commission 2015a, 11). Cramme, Meyer, and Ritzen (2013) state therefore that Europe should not be considered to be the new locus of government, but instead it should provide institutions to support the individual EU member states’ reform efforts. In order to increase public support for the EU, the authors urge pro-European reformers to develop a new agenda that “exhibits a greater clarity about policy priorities, a sharper view of where the EU can actually add real value, and a new institutional compromise that can increase the responsiveness of democratic politics in Europe” (p. 1).

Appendix

Table 7:

Descriptive statistics, Western EU member countries, 2006–2011 (N=85,881).

Table 8:

Descriptive statistics, Eastern EU member countries, 2006–2011 (N=51,457).

Table 9:

Robustness check of standard errors of Table 2 on Euroskepticism (EUS) and negative financial expectations (NFE).

Table 10:

Robustness check: recursive bivariate probit, 2007–2011 including macro-changes, Euroskepticism (EUS) and negative financial expectations (NFE).

References

  • Anderson, C. J. 1998. “When in Doubt, Use Proxies: Attitudes towards Domestic Politics and Support for European Integration.” Comparative Political Studies 31 (5):569–601. CrossrefGoogle Scholar

  • Anderson, C. J. and K. C. Kaltenthaler. 1996. “The Dynamics of Public Opinion toward European Integration, 1973–1993.” European Journal of International Relations 2 (2):175–99. CrossrefGoogle Scholar

  • Anderson, C. J. and M. S. Reichert. 1995. “Economic Benefits and Support for Membership in the EU: A Cross-National Analysis.” Journal of Public Policy 15 (3):231–49. CrossrefGoogle Scholar

  • Armingeon, K. and B. Ceka. 2014. “The Loss of Trust in the European Union during the Great Recession since 2007: The Role of Heuristics from the National Political System.” European Union Politics 15 (1):82–107. CrossrefGoogle Scholar

  • Atkinson, A. B. 2013. “Reducing Income Inequality in Europe.” IZA Journal of European Labor Studies 2:12. Google Scholar

  • Beckfield, J. 2006. “European Integration and Income Inequality.” American Sociological Review 71:964–85. CrossrefGoogle Scholar

  • Beckfield, J. 2009. “Remapping Inequality in Europe: The Net Effect of Regional Integration on Total Income Inequality in the European Union.” International Journal of Comparative Sociology 50 (5–6):486–509. CrossrefGoogle Scholar

  • Beckmann, J., A. Belke and M. Kühl. 2011. “Global Integration of Central and Eastern European Financial Markets – The Role of Economic Sentiments.” Review of International Economics 19 (1):137–57. CrossrefGoogle Scholar

  • Braun, D. and M. Tausendpfund. 2014. “The Impact of the Euro Crisis on Citizen‘s Support for the European Union.” Journal of European Integration 36 (3):231–45. CrossrefGoogle Scholar

  • Bruter, M., 2005. Citizens of Europe? The Emergence of a Mass European Identity. Basingstoke: Palgrave Macmillan. Google Scholar

  • Bryan, M. L. and S. P. Jenkins. 2015. “Multilevel Modelling of Country Effects: A Cautionary Tale.” European Sociological Review. doi:. Crossref

  • Burgoon, B.. 2013. “Inequality and Anti-Globalization Backlash by Political Parties.” European Union Politics 14 (3):408–35. CrossrefGoogle Scholar

  • Cameron, A. C., J. B. Gelbach and D. L. Miller. 2008. “Bootstrap-based Improvements for Inference with Clustered Errors.” The Review of Economics and Statistics 90 (3):414–27. CrossrefGoogle Scholar

  • Cichowski, R. 2000. “Western Dreams, Eastern Realities. Support for the European Union in Central and Eastern Europe.” Comparative Political Studies 33 (10):1243–78. CrossrefGoogle Scholar

  • Cramme, O., A. Meyer and J. Ritzen. 2013. “A New Promise for Europe. How the Elections to the European Parliament can Stop Eurosion.” Policy Network Paper. Accessed June 2015. http://www.policy-network.net/publications/4453/A-New-Promise-for-Europe. 

  • de Vreese, C. H. and H. Boomgaarden. 2005. “Projecting EU Referendums: Fear of Immigration and Support for European Integration.” European Union Politics 6 (1):59–82. CrossrefGoogle Scholar

  • de Vries, C.E. 2013. “Ambivalent Europeans? Public Support for European Integration in East and West.” Government and Opposition 48 (3):434–61. CrossrefGoogle Scholar

  • Diez Medrano, J. 2010. “Europe’s Political Identity: Public Sphere and Public Opinion.” In European Stories: Intellectual Debates on Europe in National Contexts, edited by J. Lacroix and K. Nicolaides, 315–33. Oxford: Oxford University Press. Google Scholar

  • Eichenberg, R. C. and R. J. Dalton. 1993. “Europeans and European Community: The Dynamics of Public Support for European Integration.” International Organization 47 (4):507–34. CrossrefGoogle Scholar

  • Eichenberg, R. C. and R. J. Dalton. 2007. “Post-Maastricht Blues: The Transformation of Citizen Support for European Integration 1973–2004.” Acta Politica 42 (2–3):128–52. Google Scholar

  • European Commission. 2012. EU Budget 2012. Financial Report. Table: Operating budgetary balances, 2000–12. 

  • European Commission. 2015a. Public Opinion in the European Union, First Results. Standard Eurobarometer 83. 

  • European Commission. 2015b. Future of Europe, Summary. Special Eurobarometer 394. 

  • European Commission: Eurobarometers 65.2, 67.2, 70.1, 71.1, 71.3, 72.4, 73.4, 75.3. (2006–2011). GESIS Data Archive, Cologne.

  • Eurostat Database. 2013a. Gini Coefficient of Equivalised Disposable Income (source: SILC) (2006–2011). 

  • Eurostat Database. 2013b. Gross Domestic Product at Market Prices (2006–2011). 

  • Eurostat Database. 2013c. Harmonized Indices of Consumer Prices (HICPs) (2006–2011). 

  • Eurostat Database. 2013d. Unemployment Rates by Sex and Age Groups (2006–2011). 

  • Eurostat Database. 2015. Real Gross domestic product (GDP) growth rate – volume (2004–2014). 

  • Franklin, M. and C. Wlezien. 1997. “The Responsive Public: Issue Salience, Policy Chance and Preferences for European Unification.” Journal of Theoretical Politics 9 (3):347–63. CrossrefGoogle Scholar

  • Gabel, M. 1998a. Interests and Integration: Market Liberalization, Public Opinion, and European Union. Ann Arbor:University of Michigan Press. Google Scholar

  • Gabel, M. 1998b. “Public Support for European Integration: An Empirical Test of Five Theories.” The Journal of Politics 60 (2):333–54. CrossrefGoogle Scholar

  • Gabel, M. and H. Palmer. 1995. “Understanding Variation in Public Support for European Integration.” European Journal of Political Research 27 (1):3–19. CrossrefGoogle Scholar

  • Gabel, M. and G. D. Whitten. 1997. “Economic Conditions, Economic Perceptions, and Public Support for European Integration.” Political Behavior 19 (1):81–96. CrossrefGoogle Scholar

  • Gelman, A. and J. Hill. 2009. Data Analysis Using Regression and Multilevel/Hierarchical Models, 11th printing. Cambridge:Cambridge University Press. Google Scholar

  • Gill I. S. and M. Raiser. 2012. Golden Growth. Restoring the Lustre of the European Economic Model. Washington, DC:The World Bank. Google Scholar

  • Green, W. H. 2003. Econometric Analysis, 5th ed. New Jersey:Pearson Education. Google Scholar

  • Hakhverdian, A., E. van Elsas, W. van der Brug and T. Kuhn. 2013. “Euroscepticism and Education: A Longitudinal Study of Twelve EU Member States, 1973–2010.” European Union Politics 14 (4):522–41. Google Scholar

  • Harden, J. J. 2009. “A Comparison of Methods for Estimating Linear and Generalized Linear Models with Multilevel Data.” Presentation Paper at the 26th Annual Meeting of the Society for Political Methodology, July 23–25, 2009, New Haven, CT. 

  • Herzog, A. and J. A. Tucker. 2010. “The Dynamics of Support: The Winners-Losers Gap in Attitudes toward EU Membership in Post-Communist Countries.” European Political Science Review 2 (2):235–67. CrossrefGoogle Scholar

  • Hooghe, L. and G. Marks. 2005. “Calculation, Community and Cues. Public Opinion on European Integration.” European Union Politics 6 (4):419–43. CrossrefGoogle Scholar

  • Hooghe, L. and G. Marks. 2007. “Sources of Euroscepticism.” Acta Politica 42:119–27. CrossrefGoogle Scholar

  • Kucia, M. 1999. “Public Opinion in Central Europe on EU Accession: The Czech Republic and Poland.” Journal of Common Market Studies 37 (1):143–52. Google Scholar

  • Kuhn, T., E. Van Elsas, A. Hakhverdian and W. van der Brug. 2014. “An Ever Wider Gap in an Ever Closer Union: Rising Inequalities and Euroscepticism in 12 West European Democracies, 1975–2009.” Socio-Economic Review. Advanced online publication: doi:. Crossref

  • Levy, N. and B. Phan. 2014. “The Utility of Identity: Explaining Support for the EU after the Crash.” Polity 46 (4):562–90. CrossrefGoogle Scholar

  • Loveless, M. and R. Rohrschneider. 2011. “Public Perceptions of the EU as a System of Governance.” Living Reviews in European Governance 6 (2). Accessed June 2015. http://www.livingreviews.org/lreg-2011-2. Google Scholar

  • Lubbers, M. and E. Jaspers. 2011. “A Longitudinal Study of Euroskepticism in the Netherlands: 2008 versus 1990.” European Union Politics 12 (1):21–40. CrossrefGoogle Scholar

  • Maddala, G. 1983. Limited-Dependent and Qualitative Variables in Econometrics. Cambridge:Cambridge University Press. Google Scholar

  • Mau, S. 2005. “Europe from the Bottom: Assessing Personal Gains and Losses and its Effects on EU Support.” Journal of Public Policy 25 (3):289–311. CrossrefGoogle Scholar

  • Mau, S., J. Mewes and N. M. Schöneck. 2012. “What Determines Subjective Socio-economic Insecurity? Context and Class in Comparative Perspective.” Socio-Economic Review 10 (4):655–82. Google Scholar

  • McLaren, L. 2002. “Public Support for the European Union: Cost/Benefit Analysis or Perceived Cultural Threat?”The Journal of Politics64 (2):551–66. CrossrefGoogle Scholar

  • McLaren, L. 2007. “Explaining Mass-Level Euroskepticism: Identity, Interests, and Institutional Distrust.” Acta Politica 42 (2–3):233–51. CrossrefGoogle Scholar

  • OECD. 2008. Growing Unequal? Income Distribution and Poverty in OECD Countries. Paris:OECD Publishing. Google Scholar

  • OECD. 2011. Divided We Stand: Why Inequality Keeps Rising. Paris:OECD Publishing. Google Scholar

  • Pew Research Center. 2015. Faith in European Project Reviving. Accessed July 2015. http://www.pewglobal.org/files/2015/06/Pew-Research-Center-European-Union-Report-FINAL-June-2-20151.pdf. 

  • Piketty, T. 2014. Capital in the Twenty-First Century.Cambridge, MA:Harvard University Press. Google Scholar

  • Ritzen, J. and K. F. Zimmermann. 2014. “A Vibrant European Labor Market with Full Employment.”IZA Journal of European Labor Studies 3:10. Google Scholar

  • Rohrschneider, R. and S. Whitefield. 2004. “Support for Foreign Ownership and Integration in Eastern Europe: Economic Interests, Ideological Commitments, and Democratic Contexts.” Comparative Political Studies 37 (3):313–39. CrossrefGoogle Scholar

  • Rotte, R. and K. F. Zimmermann. 1998. “Fiscal Restraint and the Political Economy of EMU.” Public Choice 94 (3):385–406. CrossrefGoogle Scholar

  • Sanchez-Cuenca, I. 2000. “The Political Base of Support for European Integration.” European Union Politics 1 (2):147–71. CrossrefGoogle Scholar

  • Serricchio, F., M. Tsakatika and L. Quaglia. 2013. “Euroscepticism and the Global Financial Crisis.” Journal of Common Market Studies 51 (1):51–64. Google Scholar

  • Summers, L. H. 2013. “IMF Fourteenth Annual Research Conference in Honor of Stanley Fischer.” Washington, DC, November 8, 2013. 

  • Stegmueller, D. 2013. “How Many Countries for Multilevel Modeling? A Comparison of Frequentist and Bayesian Approaches.” American Journal of Political Science 57 (3):748–61. CrossrefGoogle Scholar

  • Wilde, J. 2000. “Identification of Multiple Equation Probit Models with Endogenous Dummy Regressors.” Economics Letters 69:309–12. CrossrefGoogle Scholar

Footnotes

  • 1

    “Euroskepticism” does not necessarily refer to the European currency as some people infer. Hence, it would probably be better to call it “EU skepticism.” However, an established literature uses the term “Euroskepticism,” which we follow in this article. Although EU membership does not necessarily imply EMU membership, increasing problems with the governance of the Euro area may have also been a concern of many participants in the study. 

  • 2

    Loveless and Rohrschneider (2011) provide an extensive and informative literature review with regard to Euroskepticism explanatory approaches on which this section is partly based. 

  • 3

    We refrain here from the explanatory approach of social location that is related to post-materialism, cognitive mobilization or religion, because we do not consider it essential to our argumentation. For an overview, see Loveless and Rohrschneider (2011). 

About the article

Published Online: 2015-12-05

Published in Print: 2016-04-01


Citation Information: The B.E. Journal of Economic Analysis & Policy, ISSN (Online) 1935-1682, ISSN (Print) 2194-6108, DOI: https://doi.org/10.1515/bejeap-2015-0052.

Export Citation

©2016 by De Gruyter. Copyright Clearance Center

Citing Articles

Here you can find all Crossref-listed publications in which this article is cited. If you would like to receive automatic email messages as soon as this article is cited in other publications, simply activate the “Citation Alert” on the top of this page.

Comments (0)

Please log in or register to comment.
Log in