(Image source: Christian Lue on Unsplash)

Contents:

• Introduction
• Literature
  • Euroscepticism
  • European Elections
  • Corruption
• Data and Methodology
• Results
• Discussion
• Conclusion
• References
• Appendices

Introduction

Euroscepticism is a phenomenon that has been embedded into the framework of the European Union from the foundation of the European Coal and Steel Community in 1951. The early years of integration, revolving around the expansion of competencies of the European Economic Community, were characterized by French resistance to encroachment on their sovereignty, which culminated in the ‘empty chair’ crisis in which French President Charles de Gaulle pulled his ministers from their positions within European institutions as a protest. While this move was ultimately unsuccessful from the standpoint of de Gaulle’s ultimate vision of a Europe of strictly sovereign states, the Eurosceptic movement has been a constant feature at each stage of the European integration process. This movement has made itself a continual presence especially within the institution of the European Parliament, where MEPs have gotten elected on Eurosceptic tickets and Eurosceptic groups have formed within the EP on both the left and right ends of the political ideology spectrum. The 2010s saw a sharp growth in this phenomenon; while centrist Europhile groups continued to form the majority in the EP, Eurosceptic groups saw great success in France (where the far-right Eurosceptic National Front party won a plurality of the vote in 2014 and again in 2019 as the National Rally), Belgium, the Netherlands, and especially Britain, where Nigel Farage led the Brexit movement from UKIP’s 2014 victory and the Brexit Party’s 2019 victory.

To what extent can the rise of Eurosceptic movements within specific European member-states be attributed to protest against governance failures in those same states? This paper examines this question by studying the effect of the state’s ability to control corruption on the proportion of allotted EP seats in that state elected from Eurosceptic groups.  This paper will unfold in the following manner: first, the literature on Euroscepticism, quality of governance, and European supranational elections will be examined; second, the paper will discuss the data and methodology, as well as outlining the hypothesis derived from the literature; third, this paper will discuss the results of the statistical analysis and provide some avenues for future research; finally, this paper will conclude that higher levels of corruption are indicative of larger proportions of Eurosceptic MEPs.

Literature

Euroscepticism

Euroscepticism is a phenomenon that has been studied extensively without scholars ever coming to a consensus on how it comes about or what it means. Part of this is because the term “Eurosceptic” encompasses a broad set of opinions and policies regarding European integration and institutions that runs along the entirety of the left-right ideological spectrum. The term “Eurosceptic” emerged from the British press in the debate over the Single European Act in the mid-1980s and was incorporated at the elite level for the first time in UK Prime Minister Margaret Thatcher’s 1988 speech at Bruges (Leconte, 2010: 3). British reticence toward increased integration of the EU was the initial focus of study using this term, although the failure of the Danish accession to the Maastricht Treaty (and the near-failure of the French accession to the same) brought the level of analysis to the European level as a whole. Generally speaking, the term “Eurosceptic” refers to any critique of the direction of European integration; this overly broad definition has led scholars to build typologies of Eurosceptic movements in order to better differentiate, for example, opposition to certain aspects of the Common Agricultural Policy from opposition to EU membership at all.

One major division in the study of Eurosceptic groups and movements is that between the “Sussex school” and the “North Carolina school” of scholarship. The Sussex school uses the “hard/soft” division in types of Euroscepticism, where the divide is that between “principled objection” and “qualified objection” – objection to the European project as a whole or objection to the way in which European integration has been accomplished to date (Szczerbiak and Taggert, 2008). The North Carolina school studies Euroscepticism in terms of parties and their effects on the opinions of voters, finding that “party positions do influence electorate opinion, but that this effect varies with levels of disagreement among parties, party unity, issue salience, and party attachment” (Ray, 2003: 978). Other scholars have attempted more finely-grained typologies. Petr Kopecky and Cas Mudde (2002) distinguish between ‘diffuse’ and ‘specific’ support for the European project, where the former refers to support for the ideas of European integration and the latter to support specifically for the European Union as it is formulated at the time of study. Richard Haesly (2001) examines attachment to the European project in Scotland and Wales by differentiating between Eurosceptics (who are opposed to most of the integration process), Europhiles (who are in favour of a pan-European integrated identity) and instrumental Europeans (who see attachment to the European project as a way of differentiating themselves from the majority group within the member-state) (Haesly, 2001: 96). Sofia Vasilopoulou crafts a typology of Euroscepticism specifically among far-right parties that differentiates between “anti-system” parties (who are opposed to both the EU and liberal democracy, with an ambivalence toward democracy in general), “anti-liberal” parties (who are in favour of radical change within the European integration project and are critical of liberal democracy, while being in favour of democratic foundations), and “normalised” parties, who function as part of the European system and are at worst ambivalent to aspects of liberal democracy (Vasilopoulou, 2018: 46). While the broad definition of the phenomenon lends itself to large numbers of potential configurations, most studies on the subject tend to use either the Sussex or North Carolina schools as a framework for scholarship.

European Elections

The European Parliament has been a supranationally elected body since 1979. At the time, it was warned that simply allowing for an elected body would eventually lead to a ‘democratic deficit’; given that the majority of EU decisions come from the European Commission (an unelected body) and are passed largely through the European Council (another unelected body, although one comprised of officials elected to hold office at the national level) the idea is that the European Parliament is not accountable for decisions that are actually made and carried out by ‘unelected Brussels bureaucrats’ (Leconte, 2010: 55). Although there is some truth to this (administrative legislation is largely the province of the Commission and only ‘significant’ legislation is debated and voted on by the EP), there is a large role for the EP in shaping and passing a great deal of legislation. The EP has also taken on an increased role since the 2009 Lisbon Treaty, which provided for the election of the Commission President via the mechanism of the EP election. Despite this, parties on both sides of the ideological continuum use the critique that decisions in the EU are made without democratic input.

Complicating matters is the line of scholarship that examines European Parliament elections as ‘second-order elections.’ These types of elections are those where the elections are not to select a national executive or legislature; these include local elections and regional elections. They are less decisive and thus less important for voters; as such they feature lower turnouts and outcomes where “national electorates use EP elections as a ‘by-election’ to signal their pleasure or displeasure with their domestic governments” (Carrubba and Timpone, 2005: 262). The study of European Parliament elections as second-order elections was first undertaken by Karlheinz Reif and Hermann Schmitt (1980) as a response to the first 1979 elections and outlined three factors that characterize EP elections as second-order: first, turnout in EP elections is lower than in national elections; second, that government parties will lose vote share when compared to previous national elections; third, that the vote shares of small parties will increase at the expense of larger parties (Reif and Schmitt, 1980; Reif, 1984). Subsequent studies of European elections have confirmed the ‘second-order election’ thesis (see Marsh, 1998; Bressanelli, 2015; Prosser, 2016; Boomgaarden, et al 2016; and Schmitt and Teperoglou, 2015). Given this, Eurosceptic parties often emerge from the far left or right and hinge their campaigns on domestic issues that are framed in terms of cultural, national, or political complaints (Leconte, 2010: 67). Other strains of research, however, find that despite the second-order election theory explaining much of the variation in turnout and outcome there are truly European issues that drive voters to the polls as well. Richard Flickinger and Donley Studlar (2007), for example, find that inclusion of a measure indicating the relative strength of the agricultural industry within a member-state had a statistically significant impact on turnout rates in the 2004 EP election (Flickinger and Studlar, 2007: 396). The importance of certain core competencies of the EU, then, can have an effect on electoral outcomes at the supranational level.

Corruption

Corruption, in terms of the literature on governance, is recognized as a problem that has existed throughout much of history. The standard definition used in much of the corruption literature is that corruption is “the misuse of public office for private gain” (Treisman, 2000: 399). This definition is intentionally broad and encompasses both ‘petty’ corruption (bribes, campaign finance violations, and other regulatory matters) and ‘grand’ corruption, which Susan Rose-Ackerman (1999) defines as “collusion among the highest levels of government that involves major public sector projects, procurement, and large financial benefits among high-level public and private elites.” Bo Rothstein (2011) distills this down to the idea that corruption is a factor in the overall quality of a state’s governance, and that quality here can be summed up succinctly as “impartiality” – the idea that everyone is treated the same when it comes to the ability to access public goods. Corruption in this framework is really partiality, where an individual or a group is treated differently either because they are paying for that treatment (such as the payment of bribes), because the regulatory mechanisms have been captured by private interests, or because jobs and goods are being channeled toward a specific group in exchange for their votes. Carolyn Warner (2007) notes that much of the corruption that is detected and discussed within the European Union comes about for the “purposes of funding political parties and campaigns” (Warner, 2007: 16). Corruption has been studied in terms of its effect on turnout in elections at the national level (Dahlberg and Solevid, 2016; Griesshaber and Geys, 2012; Stockemer, 2013; Stockemer et al. 2013; Sundstrom and Stockemer, 2015) and the sub-national level (Carlson and Reed, 2013; Lacombe et al, 2016; Stockemer and Calca, 2013) with mixed results. The study of the connection between national levels of corruption and electoral results for Eurosceptic groups is one that has not been delved into deeply.

Data and Methodology

In order to analyze the connection between perceived levels of corruption and outcomes for Eurosceptic parties, this study uses a dataset comprising aggregate-level results by member-state for the three European Parliament elections since 2009. Official electoral results are gathered from the European Parliament website; other variable measurements are taken from World Bank data, provided through their web portal. There are 27 member-states represented in the 2009 EP elections; the integration of Croatia into the EU in 2013 gives the 2014 and 2019 electoral results 28 member-states each, resulting in a sample size of 83 states.  

The dependent variable here is the proportion of a state’s district magnitude that is occupied by Eurosceptic MEPs. In order to differentiate Eurosceptic MEPs from ordinary or Europhile MEPs, and to better encompass the broad nature of Eurosceptic policy, the decision to place a MEP into one camp or another is accomplished through the use of European Parliament affiliation groups. The European Parliament is, after the 2019 election, divided into seven groups, each claiming a particular niche of supporters. Of these seven, three encompass political parties that embrace Euroscepticism on the continuum from soft to hard, as per Szczerbiak and Taggert. These are European Conservatives and Reformists (originally founded by British Prime Minister David Cameron in 2009 as the Movement for European Reform), Identity and Democracy (a right to far-right grouping that includes parties such as France’s National Rally and the Freedom Party of Austria that was founded as the Movement for a Europe of Nations and Freedom in 2014), and the European United Left/Nordic Green Left (a left-wing democratic socialist grouping with communist undertones founded as European United Left in 1995 whose critique of the EU’s direction and vision pairs with their critique of free trade and capitalism). Non-attached members are treated according to where their party grouped in later elections; for example, both the Freedom Party of Austria and Belgium’s Vlaams Belang won seats in 2009 as non-attached members and later joined what is currently Identity and Democracy after 2014. A complete list of parties included in the rubric of Eurosceptic is included in Appendix A.

The particular measurement of the dependent variable as a proportion makes a standard ordinary least squares regression analysis difficult, as values that are measured with the extreme values of 0 and 1 often feature heteroskedasticity and a pronounced distribution skew. As such, this study makes use of beta regression, which assumes that the dependent variable is beta-distributed (that is, it takes a value in the range (0,1). The betareg package in R was created to take advantage of this type of regression; it produces regression objects that are as close as possible to those from generalized logit models and provides coefficients that are “interpretable in terms of the mean of the response,” meaning that they represent a change above or below the background mean (Ferrari and Cribari-Neto, 2004). The regression model uses a logit link to fit values, so that the results always fall between 0 and 1; the coefficients can be exponentiated to provide odds ratios of the explanatory variable having an effect on the dependent variable (Cribari-Neto and Zeileis, 2010). Given the presence of zeroes in the dependent variable, however, a transformation must be applied to the dependent variable (Smithson and Verkulien, 2006) using the following formula:

Y(n – 1) + s / n

Where Y is the dependent variable, n is the total number of observations, and s is a constant between 0 and 1; the value used here is 0.5, which has been previously identified as a “reasonable choice” (Smithson and Verkulien, 2006).

The explanatory variable here is the state’s perceived ability to control corruption, measured here using the Control of Corruption indicator from the World Bank’s Worldwide Governance Indicator set. This variable is a standardized measure that ranges from 2.5 to -2.5 with a mean of 0, where higher scores indicate a greater ability on the part of the state to control corruption within its borders. Given the findings previously uncovered through the literature, this paper offers the hypothesis that corruption will have a significant effect on the proportion of Eurosceptic MEPs. Specifically, the hypothesis is: States with higher levels of corruption will feature higher proportions of Eurosceptic MEPs in the slate of members being sent to the European Parliament. The second-order nature of EP elections will cause voters to be more willing to cast “protest votes” against established political groups at the national level, given previous research, and this effect will be amplified in states where corruption levels are higher.

To control for other possible covariates in the Eurosceptic vote, a number of secondary explanatory variables are used. To control for the level of development, the state’s GDP per capita in the year prior to the election is used; this variable is log-transformed to promote linearity and aid in interpretation. The percentage of a state’s population living in urban settings is used as a proxy for the strength of the agricultural sector, where it is assumed that less urban states feature larger agricultural concerns, and thus less Eurosceptic electoral successes. This is used to model the effect of European-issue voting, as per Flickinger and Studlar (2007).

In order to model the effect of democratic deficit at the supranational level, the relative representation index is used. The relative representation index is derived from the following formula:

(Xµ/Yµ)/(ƩXµ/ƩYµ)

Where X is the number of seats for state µ and Y is the population for state µ (Yamamura 2011; Kawaura 2003). This index is a measure of how strong a given state’s representation in the European Parliament is relative to the strength of other state’s representation. Given the Eurosceptic vote’s protest role in terms of second-order elections, it’s expected that higher Eurosceptic representation should occur in states whose overall representation is lower than other states. A table providing the index scores for each state is shown in Appendix B. Finally, the level of political engagement in the state is measured using the turnout, derived from official statistics provided by the European Parliament.

Results

Table 1 provides summary statistics of the variables involved in the analysis. The results of the beta regression are provided in Table 2.

Table 1

Variable	Min	Max	Median	Mean	Standard Deviation
Eurosceptic Seats	0	0.64	0.19	0.21	0.17
Control of Corruption	-0.27	2.4	0.98	1.03	0.8
GDP Per Capita	7261	116639	29059	35659	22813.99
Urban Percentage	52.21	98	71.62	73.06	12.62
Relative Representation Index	0.74	9.54	1.43	2.05	1.87
Turnout	13.05	90.76	43.81	46.1	17.93

Table 2

Variable	Coefficient (Standard Error)	Odds Ratio
Control of Corruption	-0.65 (0.24) **	0.52
Log (GDP Per Capita)	0.84 (0.34) *	2.31
Urban Percentage	0.03 (0.01) **	1.03
Relative Representation Index	-0.3 (0.06) ***	0.74
Turnout	-0.01 (0.008)	0.99
Constant	-10.31 (3.25) **	—
Pseudo R2	0.42	—
Log-Likelihood	68.4 on 7 df	—

Corruption is a significant and fairly strong factor in the success of Eurosceptic parties in recent European Parliament elections. Greater perceived control over corruption on the part of the state is correlated to lower proportions of Eurosceptic MEPs being sent to Brussels from each election. A visualization of the effect of the variable is shown in Figure 1.

Figure 1

As shown in Figure 1, moving from 0 to 1 in the Control of Corruption index drops the predicted proportion of Eurosceptic MEPs from just under 40% to approximately 20%. Moving from 1 to 2 has a slightly more muted effect, however, with a drop of only 10%. There is some evidence, then, that corruption drives Eurosceptic electoral success when the problem is more widespread within the country than otherwise.

  The level of economic development has a strong, significant effect as well; states featuring higher levels of GDP per capita feature higher proportions of elected Eurosceptic MEPs. The development factor has the highest odds of producing greater proportions of Eurosceptic MEPs of any of the variables examined. Figure 2 shows the effect of the variable at different levels. The strongest effect occurs between $20,000 and $40,000, which raises the predicted proportion of Eurosceptic MEPs from around 15% to just under 30%. The effect of development tapers off after the $40,000 mark; moving from $40,000 to $60,000 increases the predicted proportion of Eurosceptic MEPs by only slightly less than 10%.

Figure 2

The effect of the urban population variable, shown in Figure 3, has a positive significant effect on the predicted proportion of Eurosceptic MEPs. The most rural member-states have a low predicted amount, around 15% of the total slate of MEPs. At the other, more highly urbanized end, this predicted amount rises to nearly 40%. Interestingly, the odds ratio of this is quite slight, at only 1.03. The discrepancy between the odds ratio and the effect is likely due to the presence of Belgium acting as an outlier; Belgium’s urbanization rate is the highest, and it also features strong electoral outcomes for Eurosceptic MEPs, especially from the Vlaams Belang party. The results of a Cook’s Distance plot (included as Appendix C) show that the only statistical outlier in the dataset is Poland 2019, however; it is unclear whether the discrepancy here is an aberration or if it is an amplification of an existing but typically much more muted effect.       

Figure 3

The relative representation index, shown in Figure 4, also shows a significant effect on the proportion of Eurosceptic MEPs. Member-states with low relative representation in the European Parliament are also states more likely to send Eurosceptic MEPs there. Moving from the lowest measured value to 2 lowers the proportion of Eurosceptic MEPs by 10%; this effect is fairly consistent across all recorded values, although the sparsity of observations at the highest end reduces confidence in this consistency somewhat.

Figure 4

The effect of turnout, the proxy for a baseline level of political participation, is not significant although the sign and average marginal effect shows that higher levels of participation trend toward lower proportions of Eurosceptic MEPs. Figure 5 shows a 15% drop in Eurosceptic MEPs when moving from one end of the range to the other, but the confidence intervals reflect the statistical uncertainty of this effect.

Figure 5

Discussion

Corruption has a clear effect on the election of Eurosceptic MEPs, with lower levels of corruption producing lower proportions of Eurosceptic MEPs elected. This shows evidence for the hypothesis that voters in European Parliament elections see supranational elections as a chance to express their anger at nationally governing parties by voting for parties whose views tend more toward critique of the political status quo at either end of the left-right continuum. Corruption, as noted previously, is a phenomenon that is used as weight in these critiques, and voters in states with less control over corruption turn out to vote for parties offering these critiques in greater numbers in this analysis. This fits with the theory of second-order elections: with less at stake, voters feel freer to vote for parties that would be considered more to the fringe in national elections but are addressing a critique toward the elected mainstream parties that hold power within the state itself. In essence, the larger presence of corruption prompts voters to go to the polls to “throw the bums out” in a symbolic fashion that carries less direct political risk.

The additional control variables are also of interest. The development measurement shows some evidence for the idea that less economically developed states don’t send Eurosceptic MEPs to Brussels at a particularly high rate. The non-linear nature of the result, however, shows that the effect of economic development plateaus to an extent. The other variables provide much more straightforward linear effects. The effect of urbanization is interesting when considering the competencies of the EU; given that more urbanized member-states feature a higher proportion of Eurosceptic MEPs, it may well be that supranational agricultural legislation such as the Common Agricultural Policy is well-received by more agriculturally-focused states, prompting them to embrace Euroscepticism to a lesser extent than more urbanized states. Future work would be necessary to examine the link between agricultural policy, rural voting patterns, and the electoral success of Eurosceptic groups.            

Finally, the electoral variables show mixed results. The effect of turnout levels on Eurosceptic electoral success is linear but non-significant. It may be that using turnout to capture the level of political participation in a state’s electorate may be too broad a measure; it is also possible that the second-order nature of the election makes the turnout an inappropriate measure for participation levels. The linear nature and direction of the result, however, provide some basis for further research into the link between lower levels of political participation and higher support for Eurosceptic groups and movements. Representation, on the other hand, seems to matter to a significant degree. States that are not as represented in the EP as others are states that feature higher electoral results for Eurosceptic groups. This speaks to the idea that Eurosceptic groups are channeling disillusion with the way that supranational institutions are formulated, and that voters who turn out in greater numbers for such groups are reacting in part to a feeling of “national frustration” in having less of a say in the legislative procedure of the EU than other states when placed on an equal basis.

Conclusion

The phenomenon of corruption is one that has had an effect on politics for a long time, electoral or otherwise. Contemporary studies of corruption with regard to elections have often focused on turnout, while fewer have examined the effect of corruption on the outcomes of elections in terms of their ideological or categorical groupings. This study shows that corruption plays a role in the strength of the Eurosceptic vote at the supranational level. This is in itself a function of the second-order nature of European Parliament elections; the lower stakes of the election causes voters to feel more comfortable casting protest votes; this appears to be especially the case in states where corruption is a more endemic problem than elsewhere. This finding fits well within second-order theory, which holds that governing parties tend to lose vote share at the expense of smaller parties. Corruption seems to be a driving factor behind this effect, at least in terms of high-corruption states, and controlling for other potential covariates that may be driving the Eurosceptic vote.

The Eurosceptic movement is one that has been a presence in the European Union since the beginning of the community, and while it has reached an apparent apex with the exit of the United Kingdom from the EU it is likely to remain a political force to a varying extent across many of the remaining member-states. Future research should endeavour to examine the links between this movement and dynamics of national politics that spill over to become issues on the supranational level. Corruption is one such national dynamic, and one that cuts across ideological divides to function as an issue of concern that can be used as a campaign issue by all parties. Future research in this avenue would benefit from a close examination of each party and their use of rhetoric concerning corruption at both the national and supranational level.

References

Boomgaarden, H. G., D. Johann and S. Kritzinger. Voting at National Versus European Elections: An Individual Test of the Second Order Paradigm for the 2014 European Parliament Elections. Politics and Governance 4 (1): 130-145.

Bressanelli, E. (2015). The European Elections of May 2014: Still Second-Order Elections? Italian Politics: The Year of the Bulldozer 30: 83-102.

Carlson, M. and S. R. Reed (2013). Election Law Violations as Campaign Effort: Turnout in Japan’s House of Councillors Elections. Democratization 20 (7): 1243-1267.

Carruba, C. and R. J. Timpone (2005). Explaining Vote Switching Across First- and Second-Order Elections: Evidence from Europe. Comparative Political Studies 38 (3): 260-281.

Cribari-Neto, F. and A. Zeileis (2010). Beta Regression in R. Journal of Statistical Software 34 (2): 1-24.

Dahlberg, S. and M. Solevid (2016). Does Corruption Suppress Voter Turnout? Journal of Elections, Public Opinion and Parties 26 (4): 489-510.

Ferrari, S. and F. Cribari-Neto (2004). Beta Regression for Modelling Rates and Proportions. Journal of Applied Statistics 31 (7): 799-815.

Flickinger, R. and D. T. Studlar (2007). One Europe, Many Electorates? Models of Turnout in European Parliament Elections After 2004. Comparative Political Studies 40 (4): 383-404.

Griesshaber, N. and B. Geys (2012). Civic Engagement and Corruption in 20 European Democracies. European Societies 14 (1): 57-81.

Haesly, R. (2001). Euroskeptics, Europhiles and Instrumental Europeans: European Attachment in Scotland and Wales. European Union Politics 2 (1): 81-102.

Kaufmann, D., A. Kraay and M. Mastruzzi (2010). The Worldwide Governance Indicators: Methodology and Analytical Issues. Hague Journal on the Rule of Law 3: 220-246.

Kawaura, A. (2003). Public Resource Allocation and Electoral Systems in the US and Japan. Public Choice 115: 63-81.

Kopecky, P. and C. Mudde (2002). The Two Sides of Euroscepticism: Party Positions on European Integration in East Central Europe. European Union Politics 3 (3): 297-326.

Lacombe, D., R. Coats, W. Shughart and G. Karakhan (2016). Corruption and Voter Turnout: A Spatial Econometric Approach. Journal of Regional Analysis & Policy 46 (2): 168-185.

Lectonte, C. (2010). Understanding Euroscepticism. Hampshire: Palgrave Macmillan.

Marsh, M. (1998). Testing the Second-Order Election Model after Four European Elections. British Journal of Political Science 28: 591-607.

Prosser, C. (2016). Second Order Electoral Rules and National Party Systems: The Duvergerian Effects of European Parliament Elections. European Union Politics 17 (3): 366-386.

Ray, L. (2003). When Parties Matter: The Conditional Influence of Party Positions on Voter Opinions about European Integration. Journal of Politics 65 (4): 978-1004.

Reif, K. and H. Schmitt (1980). Nine Second-Order Elections: A Conceptual Framework for the Analysis of European Election Results. European Journal of Political Research 8 (1): 3-44.

Reif, K. (1984). National Electoral Cycles and European Elections. Electoral Studies 3: 244-255.

Rose-Ackerman, S. (1999). Corruption and Government: Causes, Consequences and Reform. Chicago: University of Chicago Press.

Rothstein, B. (2011). The Quality of Government: Corruption, Social Trust, and Inequality in International Perspective. Chicago: University of Chicago Press.

Schmitt, H. and E. Teperoglou (2015). The 2014 European Parliament Elections in Southern Europe: Second-Order or Critical Elections? South European Society and Politics: The 2014 European Parliament Elections in Southern Europe 20 (3): 287-309.

Smithson, M. and J. Verkulien (2006). A Better Lemon-Squeezer? Maximum-Likelihood Regression with Beta-Distributed Dependent Variables. Psychological Methods 11 (1): 54-71.

Stockemer, D. (2013). Corruption and Turnout in Presidential Elections: A Macro-Level Quantitative Analysis. Politics and Policy 41 (2): 189-212.

Stockemer, D., B. LaMontagne and L. Scruggs (2013). Bribes and Ballots: The Impact of Corruption on Voter Turnout in Democracies. International Political Science Review 34 (1): 74-90.

Stockemer, D. and P. Calca (2013). Corruption and Turnout in Portugal: A Municipal Level Study. Crime, Law and Social Change 60 (5): 535-548.

Sundstrom, A. and D. Stockemer (2015). Regional Variation in Voter Turnout in Europe: The Impact of Corruption Perceptions. Electoral Studies 40: 158-170.

Szczerbiak, A. and P. Taggart. (2008). Opposing Europe? The Comparative Party Politics of Euroscepticism I: Case Studies and Country Surveys. Oxford: Oxford University Press.

Treisman, D. (2000). The Causes of Corruption: A Cross-National Study. Journal of Public Economics 76 (3): 399-457.

Vasilopoulou, S. (2018). Far Right Parties and Euroscepticism: Patterns of Opposition. New York: ECPR Press.

Warner, C. M. (2007). The Best System Money Can Buy: Corruption in the European Union. Ithica: Cornell University Press.

World Bank, World Development Indicators (2019). GDP per capita (current US$), Retrieved from https://data.worldbank.org/indicator/NY.GDP.PCAP.CD.

World Bank, World Development Indicators (2019). Urban population (% of total population), Retrieved from https://data.worldbank.org/indicator/SP.URB.TOTL.IN.ZS

Yamamura, E. (2011). Effects of Social Norms and Fractionalization on Voting Behaviour in Japan. Applied Economics 43: 1385-1398.

Appendices

APPENDIX A

List of Eurosceptic Parties:

Freedom Party of Austria (Austria)

Libertarian, Direct, Democratic (Belgium)

Vlaams Belang (Belgium)

New Flemish Alliance (Belgium)

Worker’s Party of Belgium (Belgium)

National Union Attack (Bulgaria)

Coalition: Bulgaria Without Censorship/Internal Macedonian Revolutionary Organisation-Bulgarian National Movement/Agrarian People’s Union/St. George’s Day Movement (Bulgaria)

IMRO-Bulgarian National Movement (Bulgaria)

Croatian Sovereignist Coalition (Croatia)

Human Shield (Croatia)

Progressive Party of the Working People (Cyprus)

Civic Democratic Party (Czech Republic)

Communist Party of Bohemia and Moravia (Czech Republic)

Free Citizen’s Party (Czech Republic)

People’s Movement Against The EU (Denmark)

Danish People’s Party (Denmark)

Conservative People’s Party of Estonia (Estonia)

The Left Alliance (Finland)

Finns Party (Finland)

Left Front (France)

Alliance of the Overseas (France)

Libertas France (France)

National Front (France) (Renamed National Rally for 2019 elections)

Unbowed France (France)

The Left (Germany)

Alternative For Germany (Germany)

Family Party of Germany (Germany)

Human Environment Animal Protection (Germany)

National Democratic Party of Germany (Germany)

Communist Party of Greece (Greece)

Coalition of the Radical Left (Syriza) (Greece)

Popular Orthodox Rally (Greece)

Independent Greeks-National Patriotic Alliance (Greece)

Golden Dawn (Greece)

Greek Solution (Greece)

Hungarian Democratic Forum (Hungary)

Movement for a Better Hungary (Jobbik) (Hungary)

Socialist Party (Ireland)

Fianna Fail Party (Ireland)

Independents 4 Change (Ireland)

Northern League (Italy)

The Other Europe (Italy)

Five Star Movement (Italy)

Brothers of Italy (Italy)

For Fatherland and Freedom (Latvia)

Harmony Centre (Latvia)

Union of Greens and Farmers (Latvia)

National Alliance (Latvia)

Electoral Action of Poles in Lithuania (Lithuania)

Order and Justice (Lithuania)

Reformed Political Party (Netherlands)

Socialist Party (Netherlands)

Party for Freedom (Netherlands)

Labour Party (Netherlands)

Forum for Democracy (Netherlands)

Law and Justice (Poland)

Congress of the New Right (Poland)

Left Bloc (Portugal)

Unitary Democratic Coalition (Portugal)

Greater Romania Party (Romania)

Slovak National Party (Slovakia)

Ordinary People (Slovakia) (*2014 only, switched to European People’s Party for 2019)

New Majority (Slovakia)

Freedom and Solidarity (Slovakia)

United Left (Spain)

Plural Left (Spain)

Decentralist Social Force Party (Spain)

The Peoples Decide (Spain)

VOX (Spain)

Coalition Ahora Republicas (Spain)

United We Can Change Europe (Spain)

Left Party (Sweden)

Sweden Democrats (Sweden)

Conservative Party (United Kingdom)

UK Independence Party (United Kingdom)

British National Party (United Kingdom)

Sinn Fein (United Kingdom)

Ulster Unionist Party (United Kingdom)

Brexit Party (United Kingdom)

APPENDIX B

State	Year	Relative Representation Index
Austria	2009	1.377
Austria	2014	1.434
Austria	2019	1.395
Belgium	2009	1.388
Belgium	2014	1.272
Belgium	2019	1.258
Bulgaria	2009	1.533
Bulgaria	2014	1.581
Bulgaria	2019	1.657
Croatia	2014	1.747
Croatia	2019	1.843
Cyprus	2009	3.75
Cyprus	2014	3.547
Cyprus	2019	3.456
Czech Republic	2009	1.432
Czech Republic	2014	1.35
Czech Republic	2019	1.353
Denmark	2009	1.599
Denmark	2014	1.565
Denmark	2019	1.537
Estonia	2009	3.032
Estonia	2014	3.078
Estonia	2019	3.111
Finland	2009	1.653
Finland	2014	1.615
Finland	2019	1.615
France	2009	0.756
France	2014	0.758
France	2019	0.757
Germany	2009	0.815
Germany	2014	0.804
Germany	2019	0.793
Greece	2009	1.342
Greece	2014	1.294
Greece	2019	1.34
Hungary	2009	1.481
Hungary	2014	1.434
Hungary	2019	1.471
Ireland	2009	1.806
Ireland	2014	1.608
Ireland	2019	1.548
Italy	2009	0.827
Italy	2014	0.819
Italy	2019	0.828
Latvia	2009	2.483
Latvia	2014	2.687
Latvia	2019	2.844
Lithuania	2009	2.535
Lithuania	2014	2.514
Lithuania	2019	2.69
Luxembourg	2009	8.307
Luxembourg	2014	7.466
Luxembourg	2019	6.77
Malta	2009	8.26
Malta	2014	9.539
Malta	2019	8.491
Netherlands	2009	1.027
Netherlands	2014	1.045
Netherlands	2019	1.033
Poland	2009	0.886
Poland	2014	0.906
Poland	2019	0.92
Portugal	2009	1.408
Portugal	2014	1.357
Portugal	2019	1.399
Romania	2009	1.086
Romania	2014	1.082
Romania	2019	1.126
Slovakia	2009	1.633
Slovakia	2014	1.623
Slovakia	2019	1.635
Slovenia	2009	2.34
Slovenia	2014	2.625
Slovenia	2019	2.643
Spain	2009	0.735
Spain	2014	0.783
Spain	2019	0.79
Sweden	2009	1.319
Sweden	2014	1.408
Sweden	2019	1.346
United Kingdom	2009	0.79
United Kingdom	2014	0.77
United Kingdom	2019	0.752

APPENDIX C