Voting for Eurosceptic Parties and Societal Polarization in the Aftermath of the European Sovereign Debt Crisis

1 Introduction

The European sovereign debt crisis has challenged European integration in the last decade: There has been a medial debate about austerity measures and economic sanctions for member states. Up to today, the role of the European Union (EU), which actively intervened in the economy of member states and the policy on interest rates by the European Central Bank and its American counterpart the Federal Reserve, is discussed controversially. That is not surprising, as in the member states these austerity measures were discussed very differently, too; accompanied by the media which accused some countries to be the “poorhouse of Europe”.^[1] This contributed to the establishment of a populist vocabulary, speaking of some countries and their citizens as “donors” and of some as “recipients” of financial aid. In two of these “receiving countries”, namely Greece and Spain, citizens started protesting against the austerity measures, their economic consequences, and the European institutions. Two new parties emerged from these protest, Syriza in Greece and Podemos in Spain. Both parties are considered to be populist, left, and highly Eurosceptic (Inglehart and Norris 2016: 44). Voting for either Podemos or Syriza in the aftermath of the financial crisis could be seen as a critique to the austerity measures and the economic hardship in both countries. But indeed, Eurosceptic parties or anti-European policies can be found in almost every European country. The prominent historic example of the UK shows that Eurosceptic parties can indicate a long-established but critical relationship of people, country, and trade unions. Sometimes, the anti-European agenda is changing over time, as the case of the German Alternative für Deutschland (AfD) demonstrates. In its beginnings, the AfD was characterised as an anti-EU party, which programmatically evolved into a right-wing party with a strong populist agenda (Lees 2018). Multiple speeches and interviews in the last two years highlighted how this includes the denial of anthropogenic climate change, and questioning of the Covid-19 policies implemented by the ‘grand coalition’ during the last two years.

The increasing influence of Eurosceptic parties in almost every European country raises the question of whether people vote for Eurosceptic parties because they perceive them as a democratic way of addressing their critique to the European Union in an institutionalized manner? If that holds true, one would expect to observe similarly high rates of Eurosceptic vote share over time and country. Or are Eurosceptic parties more successful in those countries, where anti-EU protest has already been established before, such as Spain and Greece? This would be indicated by higher voter turnout in those countries, that had to implement austerity measures in the aftermath of the financial crisis. Both questions address an ongoing scholarly debate: to what extent is voting for a party aiming to end cooperation with the EU, a reasonable criticism of the EU (Pirro, Taggart, and van Kessel 2018; Vasilopoulou 2018)? Furthermore, does European identity follow economic integration and does it prevent giving a voice to populist and Eurosceptic actors (Czech and Krakowiak-Drzewiecka 2019)? As the EU has gone through several crises since the sovereign dept crisis, these questions are still relevant today (Pirro and Taggart 2018): First, fighting the Covid-19 pandemic has created similar notions of supranational interdependencies; second, US President Donald Trump made populist argumentation and so-called fake news a daily business of Western politics; and third, Brexit set an example of a country leaving the EU, to only name a few challenges. These examples illuminate the importance of the question of how reasonable critique can be articulated in Western societies, supported by Moffits (2015) model on how populists perform crisis themselves, with reference to the argument of right-wing opportunism developed by Wren-Lewis (2016). Using the sovereign dept crisis to study that question benefits from the fact that Eurosceptic parties emerged parallel to the European integration process and as a consequence, and due to the time passed by since the crisis event happened, the data basis is sound.

I test the model by analysing survey data from the European Social Survey (ESS) (2002–2016) (Beullens et al. 2016), using data from 16 countries and controlling for time of measurement before and after the euro crisis. I choose to include Switzerland and Norway into the analysis. Both have built a strong economic alliance with the EU and profit similarly as regular member states do. The main difference is the repetitive rejection of the full membership by the citizens which led to a feeling of independence. In a financial crisis, the perceived amount of independence can be crucial.

I have conducted a trend analysis and differences-in-differences estimation of the voting behaviour to identify the effects of the crisis and of the crisis-affectedness of countries between 2002 and 2016. In doing so, I combine data from the party manifesto project to decide which of the parties asked in the ESS is supposed to be Eurosceptic and to what extent. My analyses show that the probability to vote for Eurosceptic parties depends on EU-membership duration, whilst that there is a delayed causal effect of the implementation of austerity measures as a consequence of crisis and rising levels of Euroscepticism.

This Research Note is organised as follows: In section two, I present the state of research, describe the argument and the hypotheses. Subsequently, I explain the data and methods used in section three. In section four, I present the results, before I summarize my research findings in section five.

2 Voting for Eurosceptic Parties and Polarization in Times of Crisis

Previous work on voting for Eurosceptic parties and exogenous events has focused on the supply side of Eurosceptic and populist parties, with less attention on the perceived individual insecurity arising from events such as the financial crisis. Several studies address the multiple crisis phenomena Europe has gone through the last years, its populist instrumentalization and the electoral success of Eurosceptic parties (Moffitt 2015; Pirro and Taggart 2018; Pirro, Taggart, and van Kessel 2018; Vasilopoulou 2018). But in doing so they mix up economic hardship from the Great Recession, with the unsolved assignment of European migration policy, the so-called migration crisis, with the success of the Brexit campaign. Indeed, all these events are instrumentalized by populist politicians, but the underlining mechanisms are very different. All crisis events are beneficial to such parties, since they address the corrupt elite versus the pure people conflict, and the solution to solve the crisis is to promote the general will of the people, instead of securing the status quo which is what the corrupt elite is doing (Mudde 2004: 543). However, when it comes to the political and economic consequences of the financial crisis, I argue that the phenomenon is far more complex than migration crisis or Brexit, and, due to its transnational character, must be analysed from the individual perceived threat. This Research Note contributes to the literature on the link between populism and crisis, which is often accused to be undertheorized (Moffitt 2015). Considering the work of Moffitt (2015) I close this gap by developing a theoretical model focussing on the aggregate level which shows that EU-membership and perceived insecurity is a crucial factor in explaining voting for Eurosceptic parties. This is in line with Wren-Lewis’ argument (2016: 21) that right-wing opportunists instrumentalized the “popular concern about rising government debt in order to reduce the size of the state”, which can be applied to Eurosceptic parties as well.

I assume to find a relationship between the financial crisis and rising levels of Euroscepticism. The status quo is questioned as a result of the euro crisis insecurity. Populist parties use this insecurity when blaming the European elite for its bad crisis management. Eurosceptic parties rely on the same strategy as right-wing opportunists when addressing the public concern of government deficits (Wren-Lewis 2016). This is supported by findings of Hübscher, Sattler, and Wagner (2021), who find that voter punish governments for austerity policies, while benefits from fiscal balance are weighted less (Hübscher, Sattler, and Wagner 2021). European citizens realise that the European elite is not presenting itself as a competent problem-solver to keep their status quo, and as a consequence, citizens withdraw their support and trust in European institutions (Heinisch and Schlipphak 2014). I choose an action formation mechanism that is based on rational choice theory. I assume that it is rational to restore security in terms of stability of expectations and status quo that were challenged by the crisis and the lacking problem-solving competence of EU elites. The rationalist way of expressing this is through voting for Eurosceptic parties in national elections, but not in European Parliament elections (Schäfer 2006). The rational voter anticipates that through their Eurosceptic vote on the national level, they will influence the level of European integration, that is the decision of ‘More Europe or not?’ Whereas in European Parliament elections the question is ‘How are certain European policies developed?’ (Schäfer 2006: 353).

This is in line with the so-called vulnerability argument by Heinisch and Schlipphak (2014). This argument states that European citizens are aware of the economic consequences of the financial crisis and recognise the failure of the European institutions to act as problem-solvers (Heinisch and Schlipphak 2014: 183). The influence of European institutions in areas traditionally in the realm of national politics, for example the deduction of retirement benefits, has been visible. As a consequence, citizens have developed negative attitudes towards the EU in member states directly affected. They argue that political interventions after the crisis led to the recognition of the citizens that their member states are part of a highly integrated Europe. People reject these transnational interventions and responsibilities for other member states, which results in the desire of a nationalist backlash (Heinisch and Schlipphak 2014: 185). I develop this argument for the countries which had to implement austerity measures. As the economic consequences led to the hardest restrictions in those countries, the perceived threat of European institutions on national politics should be greater than in countries which did not implement austerity measures. This should increase the vote share for Eurosceptic parties, because voting for anti-EU parties in national elections is rational a way of expressing a refusing position towards European influence.

In hypothesis H1b, I take the argument further and state that this effect is causal. In other words, I argue that because people in the countries which have received loans have experienced the negative economic consequences of a financial crisis and the restrictions in public spending, they decide to vote for Eurosceptic parties in order to have their voices heard.

The second set of hypotheses concentrates on the relationship between membership duration and scepticism. Specifically, I expect the length of EU-membership to negatively affect Euroscepticism. In other words, the longer a state has been a member of the EU, the lower should be the share of votes for Eurosceptic parties. The idea is that citizens have arranged with the influence of the EU on national politics, such as the influence on national politics, policymaking and the everyday life of people in European—this is called permissive consensus (Eder 2000: 172; Hooghe and Marks 2009; Startin and Krouwel 2013). In those member states, which have only recently joined the union, citizens might be more sceptical about these interventions in general (McLaren 2004). In the aftermath of financial crisis, when failure of European institutions became visible, this scepticism development has speeded up. That leads to hypotheses 2a and b:

Furthermore, I expect citizens in Switzerland and Norway to have the highest probability of voting for Eurosceptic parties. Although it could be argued that if rising Euroscepticism is a problem that occurs in Europe in the aftermath of the euro crisis, it should be not found in non-full-EU-member states, this argument falls to short because these countries are not member of the EU for a good reason: There is a long tradition of protest against the EU and anti-EU-referendums. Parties in these countries usually represent the hard Eurosceptic line (Taggart and Szczerbiak 2004), but with different types of scepticism (Sundlisæter Skinner 2013: 136). According to Sundlisæter Skinner (2013: 137) a left oriented postmaterialist value based Euroscepticism is found in Norway and cultural right wing Euroscepticism is found in Switzerland, but ‘the desire to protect natural resources against foreign interference is common on both the left and the right’ (Sundlisæter Skinner 2013: 137).

Consequently, I argue that (H2c) the probability to vote for Eurosceptic parties should be the highest in non-EU countries.

To sum it up, my arguments shed light on undertheorized link between voting behaviour and societal polarization in the aftermath of the European sovereign dept crisis. I argue that the European people realised through the lens of economic insecurity that voting an Eurosceptic party could be a way to express criticism of the European elite; namely the fact that those in power did a bad job preventing economic hardship in the aftermath of European sovereign dept crisis.

3 Methods Data and Methodology

I test the hypotheses using data from the European Social Survey (2002–2016). The ESS is a nationally representative survey of individuals from the age of 15, which is conducted every two years (Beullens et al. 2016; Talving 2017).^[2] However, as the European Social Survey mainly provides individual-level data, it will be supplemented by context-level data from the European Commission (2018a, 2018b, 2018c) providing information concerning the number of loans.

For the analysis, I use data from 16 countries, which are of good composition: Three of 16 countries had to implement austerity measures between 2010 and 2015 (Spain, Greece and Ireland). Two of 16, Switzerland and Norway, are not full member states of the EU, which allows making an outside comparison. In addition, the countries differ in their length of memberships. The dataset includes four founding members: Belgium, Germany, France and the Netherlands; and at least two countries of each accession to the EU up to 2004.^[3]

4 Results

To test the relationship between crisis and scepticism, and the causal effect of the financial crisis (Hypotheses 1a and 1b), I have conducted a differences-in-differences estimation (Table 1). Euroscepticism has been on the rise between 2008 and 2012 in all countries included in the analysis, as Model 1 shows. However, the coefficient of the country variable shows that Euroscepticism has decreased in those countries that have received loans as a consequence of the crisis. A closer look at the average marginal effect shows that Euroscepticism has risen from 2008 to 2012 by about 4.6% points. In countries, that had to implement austerity measures, the probability of voting for an Eurosceptic party was 6.4% points lower than in those without austerity measures. What is more, the non-significant interaction term suggests that this is not a causal effect of the implementation of austerity measures. These results indicate that voting for Eurosceptic parties is not a reaction to the austerity measures in first place, but might be interlinked with other factors, which were not tested here and should be considered in future research.

In contrast, Model 2 (2012–2016) shows a causal effect of the year*country interaction term, which might suggest that the impact of the austerity measures is delayed. In addition, Euroscepticism has risen in all countries from 2012 to 2016, but not in those countries which received credits. But the interaction shows that Euroscepticism is rising in the aftermath of the European sovereign debt crisis in countries which had to implement austerity measures, too. When comparing 2012 and 2016, the data supports the argument of democratic reaction of the people—but presumably inverse. This leads to mixed results for the first research question. People expressed their anger about an intervention by the EU not in those countries, which received credits, but instead, in countries where populists argued against supporting member states in need and where the media campaigned against countries like Greece, Eurosceptic vote share increased. This might indicate that research question two can be affirmed, meaning that Eurosceptic parties are more successful in those countries where anti-EU protest has already established before 2008.

Based on the results, I reject hypotheses 1a and partially 1b, as Euroscepticism has increased over time in all countries (H1a), and decreased in those countries that received credits, with no causal effect found in model 1, but in model 2 (H1b).

To test whether this expresses a general trend of an increasing Eurosceptic vote share in the European Union, I have analysed the predicted probabilities of voting for an Eurosceptic party depending on the length of EU-membership (2002–2016). The results in Figure 1 suggest that Euroscepticism has, indeed, increased in all groups over time, with the exception of the founding member states, in which the share has decreased from 12.2 (2002) to 11.1 (2016)%, and also the North enlargement countries, where the levels in both, 2002 and 2016, are at 15.4%.

Figure 1:

Predicted probabilities to vote Eurosceptic by country (2002–2016).

Source: Combined data from ESS (2002–2016), predicted probabilities, own calculations.

These results indicate that voting for an Eurosceptic party becomes more likely the shorter the membership of the EU, which is in line with Hypothesis 2a. In line with the hypothesis, the founding member states have the lowest share of Eurosceptic voting, so H2b can be confirmed. As argued above, the probability is the highest for non-EU members, so H2c can be confirmed, as well. Still, it should be kept in mind that these findings are based on cross-sectional data.

5 Conclusions

The probability to vote for Eurosceptic parties depends on EU-membership duration, whilst there is no clear causal effect of the recipience of loans as a consequence of the crisis on increasing Euroscepticism. Indeed, analyses show that the crisis effect is delayed. In contrast to the first model, which concentrated on the time between 2008 and 2012, the second model indicates that Euroscepticism has risen in all countries from 2012 to 2016, but not in those countries which have received loans. However, a significant interaction term suggests that Euroscepticism is rising in the aftermath of the European sovereign debt crisis in countries which had to implement austerity measures, too. This has important implications for the study of populist parties, since the usefulness of the external shocks for populist framing of policies is not restricted to short periods after the event. Moreover, the delay shows perhaps how well the framing of certain parties worked out, well in line with the debate by Moffitt (2015) and others on the theoretical framework of populism and the shape of crisis by the populists themselves. This adds to the literature of the understanding of populist framing of such events as the Brexit or even the problem-solving competence of governments during the Covid-19-pandemic.

I expected that with a longer membership duration, Europeans are more willing to accept political influence from European institutions, which reduces insecurity that arises from crisis and, thus, the probability to vote for Eurosceptic parties. The analyses have shown that the length of EU-membership does indeed negatively affect the probability to vote for an Eurosceptic party. In other words, citizens of the countries of the 2004 EU-enlargement as well as of non-full-EU-member states, have the highest probability to vote for Eurosceptic parties. This contributes to the scientific debate on European integration by showing that integration has to happen on economic and societal level. Furthermore, my results reveal that a narrow economic link between a non-full-member and the European Union does not indicate pro-European attitudes at a societal level.

When interpreting the results, it should be kept in mind that I used cross-sectional data and conducted a trend analysis, because panel data at the European level was not available in similar quality. In addition, I relied on reported voting behaviour, which is limited by the memory of respondents, desired answering, and that voting is mandatory in some of the countries under investigation. In addition, I concentrated on the institutionalised way of articulating dissatisfaction with European politics, but the anti-Corona-protestors, for example, in Canada, stress the importance of research in the field of non-institutionalised forms of political participation. Especially as populist actors instrumentalise these protests as well.

This Research Note adds to the literature of societal polarization that the cleavage between supporters and opponents of the European ideal is perhaps deeper than expected, even when being a full member of the EU. In addition, that economic integration between the EU and countries like Switzerland and Norway does not lead to a pro-European attitude automatically. This holds true for the countries of the 2004 enlargement, in which the economy is not developing at the same pace as in countries from the previous rounds of enlargement, after joining the EU. The results indicate, that in the ongoing debate between economic integration and identity, more light should be shed on the interplay between both.

To conclude, the increasing influence of Eurosceptic parties in almost every European country raises the question whether people vote for Eurosceptic parties because they perceive them as a democratic way of addressing their critique to the European Union in an institutionalized manner? My analysis did not show similarly high rates of Eurosceptic vote share over time and country. But I found higher voter turnout in those countries, that had to implement austerity measures in the aftermath of the financial crisis. A possible explanation for that is: Eurosceptic parties are more successful in those countries, where anti-EU protest has already established before, such as Spain and Greece. To shed more light on this argument further studies should address individual level explanations for anti-EU-attitudes.

Linking the results to the first research question, it must be differentiated between the articulation of anger about European crisis politics and making use of an Eurosceptic vote to address that. And the perspective of those citizens in the countries providing loans. In those countries, a narrative focussing on the fact that countries have failed to follow the financial rules of the EU before the crisis has led to an Eurosceptic shift. In other words, the shift was not driven by a lack of sovereignty but by the populist narrative that the EU is not a competent problem solver if it does not succeed in keeping the wealthy states wealthy. This contributes to the debate of European integration and its link to populism, which gained even more importance after Brexit—and the subsequent controversy about a more optimistic or pessimistic outlook on the European integration process and the division of power within the EU (Czech and Krakowiak-Drzewiecka 2019). My results are in support of the pessimistic view of European integration as an undeveloped societal task, which made it too easy for populists to jump in and instrumentalize the conflict of power division within the EU.