How the European debt crisis reshaped national political space: The case of Greece

Abstract

Where some authors saw a limited impact of Europeanisation on national party politics, others proposed that in addition to the pre-existing economic left-right dimension a separate European Union dimension structures the national political space. This article looks at the Greek bail-out during the European sovereign debt crisis to examine how Europeanisation can change the national political space. The bail-out came with memoranda that set the main lines of Greek economic policy for the coming years. Accepting these policies was connected with remaining in the Eurozone. This restructured the political space: the economic and European integration form one dimension. A second relevant dimension focuses on cultural issues. The economic/European dimension is a stronger predictor of vote choice than the cultural dimension.

Keywords

Dimensions economic crisis Europeanisation Greece party politics voting behaviour

Introduction

The extent to which European integration has shaped the national political space is a matter of academic debate: Mair (2000) has argued that the impact of European integration on national party politics is only limited. Other authors, most prominently Kriesi et al. (2008), see the formation of a European integration dimension. This dimension exists in addition to the left-right dimension and its core debate is the issue of Europeanisation, the process by which individuals and institutions respond to the altered conditions generated by European integration (Ladrech, 2002). Specifically, it concerns the Europeanisation of the national political space, the process by which the positions towards European integration structure how voters and parties think about other political issues. This article looks at an extreme case: the bail-out of Greece by the European Union (EU).

This article examines a transformational moment in Greek and EU politics brought on by the biggest economic crisis since the 1930s. It offers an extreme case to explore the effect of European integration on the national political space: the fierceness of the crisis and the complete dependency of Greece on the bail-out deal with its European partners made European integration a central issue on the Greek political agenda. Economic issues became intertwined with the question of continued Eurozone membership. In the 2012 Greek elections, supporting Eurozone membership implied committing to a set of economic policies, and rejecting these policies also meant rejecting Eurozone membership. In this way, European integration may have become the issue that structured the views of voters and parties on the most important economic questions. In summary, this article sets out to identify the process of Europeanisation by determining to what extent the question of European integration structures voter and party policy positions in post-crisis Greece.

The political results of the Greek sovereign debt crisis contrast the existing wisdom concerning the effect of Europeanisation on national politics as proposed by Kriesi et al. (2008). In their view the question of European integration is part of a new dimension that also concerns cultural issues such as immigration but is separate from the economic left-right dimension. They believe economic integration has reinforced this dimension because of stronger economic competition (Kriesi et al., 2008). The Greek case offers us a different mechanism, whereby voter and party positions towards the EU structure the economic dimension as the result of the expanded European control over economic policy.

The relevance of this study does not end at the Greek borders despite the fact that Greece is an extreme case for finding an effect of the European sovereign debt crisis on a European party system (Gerring, 2007). Similar patterns may be visible in other European countries that were bailed out. Moreover, our argument does not just concern crisis solutions, but rather the growing influence of the EU over budgetary policy. This affects all Eurozone countries, including Northern countries that have pressed for stronger budgetary guidelines. There is evidence that a similar pro/anti-EU dimension has structured party positions on economic policies in the Dutch 2012 elections (Otjes, 2015).

Left, Right and Europeanisation

Our key argument is that the processes set in motion during the on-going financial crisis have resulted in a situation in which the political issue of European integration structures the way voters and parties understand the political space in post-crisis Greece. The voter space is structured by dimensions that capture consistent differences in opinion among citizens (Kriesi et al., 2008): citizens that agree on some political questions (e.g. lowering taxes) also agree on other issues (e.g. lower immigration). The number of dimensions needed to navigate the voter space is constrained by the voters’ cognitive limits and their capacity to process information (Hinich and Munger, 1996). Like voters, party positions may also consistently vary and these differences could also be reflected in dimensions. The voter and party spaces do not necessarily need to have the same underlying structure (Van der Brug and Van Spanje, 2009).

Which dimensions structure the voter and the party space is an oft-debated question. The main issue has focused on whether a single dimension, often the general left-right, suffices to understand voting behaviour. On the one hand, the high complexity of politics forces voters to find shortcuts in the form of attitudinal cues (Feldman and Conover, 1983). A robust and enduring heuristic is the left-right dimension, which according to Fuchs and Klingemann (1989) is the predominant device for summarising the ideological position both of individual citizens and political parties. The left-right dimension is the organising element of the shared political consciousness between mass publics and elites (Laponce, 1970): a ‘super-issue’ that encompasses views of voters about economic issues but also about ethics and the environment. The left-right dimension has shown a strong integrative capacity, incorporating new issues as they enter the political agenda (Inglehart and Klingemann, 1976). On the other hand, the assumption that a single general left-right dimension structures economic and non-economic issues ‘has been shown to be implausible in numerous studies’ (Evans et al., 1996: 94). Voter positions on issues as diverse as the environment, economics and ethics do not fit into one dimension. Authors often differentiate between an economic left-right dimension and a second dimension that concerns non-economic issues (Evans et al., 1996; Kriesi et al., 2008; Lefkofridi et al., 2014; Van der Brug and Van Spanje, 2009).

According to many scholars, voter and party opinions on European integration cannot be incorporated in the left-right dimension (Gabel and Hix, 2002; Hix and Lord, 1997; Hooghe et al., 2002; Kriesi et al., 2008; Van der Brug and Van Spanje, 2009). The most prominent statement of this two-dimensional approach comes from Kriesi et al. (2008). They see the voter space as consistently two-dimensional, but argue that the nature of these dimensions has changed over time. A model with an economic dimension and a cultural dimension that concerns moral questions like abortion forms their starting point. The nature of these dimensions has changed between 1970 and the 1990s due to globalisation, immigration and European integration, which they see as one process. We focus on the effect of European integration in particular. Kriesi et al. (2008: 9) portray a development of a new structural conflict between winners and losers of globalisation. On the one hand, there are those voters and parties who favour stronger demarcation of national borders: they desire economic protectionism, see immigration as a danger and oppose their country ceding sovereignty to the EU. On the other hand, there are those parties and voters who advocate further integration. They champion economic integration into a globalised world market, believe that immigration is a positive phenomenon and want integration of the EU to deepen. Kriesi et al. (2008: 11) differentiate between an economic and a cultural dimension of the integration–demarcation divide. First, economic integration within the EU ‘reinforce[s] the classic opposition between a pro-state and a pro-market position’ (Kriesi et al., 2008: 13): the right favours integration into a European market while the left desires economic protectionism. Second, the existing cultural dimension changes in nature. It now concerns the question of political integration into the EU and openness to immigration. The new conflict concerning integration and demarcation has been integrated into the existing two-dimensional structure (Kriesi et al., 2012).

The extent to which the European sovereign debt crisis has led Kriesi and colleagues to re-evaluate their model is unclear. Kriesi et al. (2012: 20) argue that elections after the financial crisis do not provide evidence for contestation of European integration along an economic dimension. Kriesi (2014: 369–370) does note how left-wing populist parties mobilise voters against European integration in order to defend the national social welfare state model, linking European and economic issues. Finally, Kriesi and Grande (2014: 84) argue that the European debt crisis has not further politicised European integration at the mass level but rather the debate has taken an intergovernmental turn during the crisis. How we can understand the effect of the European sovereign debt crisis on the dimensions that structure the party and voter space remains an open question.

Our account builds on Kriesi et al. (2008) but also draws insights from the work of Mair (2000, 2008). He has contested the extent to which European integration forms an ‘independent base for party competition’ (Mair, 2008: 158). In his view the Europeanisation of party systems has been limited. He focuses on the question whether the issue of European integration has affected the format of the party system (that is the choices voters have) and the mechanics of the party system (the way in which parties interact with each other). Mair (2000: 36) argues that only in the United Kingdom do European citizens have ‘the potential to take Europe as a cue for choosing between governing alternatives’. Elsewhere, Eurosceptic parties played a minor role in elections and their appeal was not primarily built on their Euroscepticism. The issue of European integration does not affect the ‘mechanics’ of the party system either (Mair, 2000: 35): ‘Europe has not made for significant new alliances or enmities’.

Instead of offering citizens a choice of more or less Europe, Mair (2008: 157) describes a ‘[h]ollowing out of national party competition’ as the EU puts increasing ‘constraints on domestic decision-making’: governing parties can only choose from a stock of available policies imposed by the free market nature of the EU (Mair, 2008: 60). In all policy domains where the involvement of the EU has increased, the distance between mainstream parties’ positions has decreased (Nanou and Dorussen, 2013): only non-mainstream parties offer alternative choices.

Our contribution synthesises Mair’s (2008) argument on the constrained policy repertoire for mainstream parties and Kriesi et al.’s (2008) argument on dimensional change. Our key expectation is that in cases of sudden and far-reaching expansion of European competences over economic decision-making, the issue of more or less European integration structures economic preferences and becomes perpendicular to the cultural dimension. Kriesi et al.’s (2012, 2008) account of dimensional change is based on the assumption that favouring EU integration is unrelated to an economic agenda, allowing parties to have a pro-European integration position while placing themselves on the pro-state side of the economic spectrum. In cases, however, in which European institutions limit the economic policy repertoire to the extent that pro-European parties can only place themselves on the pro-market side of the economic spectrum (Nanou and Dorussen, 2013), European integration and economic policy become intertwined, making a pro-European, pro-state position impossible. Only Eurosceptic political parties can take diverging positions on economic issues. In such a case the polarisation over economic issues is not just strengthened by European integration, but the economic policy preferences of voters and parties may actually be structured by their position on EU integration.

Since the founding of the European Monetary Union, the direct influence of the EU over budgetary policies has increasingly limited the policy repertoire of national governments (Mair, 2008). In Eurozone countries, budgetary policies are restricted or even co-determined by the European Commission (EC) as the Stability and Growth Pact limits a country’s budget deficit to 3%. After the onset of the economic crisis, many countries were no longer able to meet this obligation. The financial markets’ trust in these countries declined. In order to prevent bankruptcy, a number of countries had to be bailed out by the Troika – a body consisting of the EC, the European Central Bank and the International Monetary Fund. The bail-out loans came with agreements (‘memoranda’) between the national government and these institutions making austerity measures, privatisation and economic reforms a requirement for cashing the loans. The goal of these measures was to regain the trust of the financial markets. The economic policies of these countries were effectively determined by these bail-out agreements. During the aftermath of the European sovereign debt crisis, the Fiscal Compact expanded the competence of the EU and the toolbox of the EC to force all Eurozone countries to keep to the Stability and Growth Pact.

As the EU increased its control over economic decision-making, room for political competition decreased and parties were forced to a take-it-or-leave-it approach. Therefore, voter and party positions about economic issues are structured by a pro-/anti-EU dimension instead of a left-right dimension. Voters opposing the 3% norm and the austerity and reform measures forced by the EC may overall oppose European control over economic decision-making. Eurosceptic parties (on the left or the right) develop a distinctive economic agenda: they oppose the reforms and austerity measures mandated by the EC and claim to defend welfare states against the ‘dictates from Brussels’. In this sense, European integration does not reinforce the economic left-right dimension but rather leads to a pro/anti-EU dynamic on economic issues. The remaining non-economic issues, such as immigration, then form a separate dimension. Given that the pro/anti-EU dimension dominates economic issues, all that remains for the general left-right dimension to tap into are cultural issues. Our expectations about the political space can be summarised as follows:

H1: Voter and party space expectation: Voter and party positions are structured by two dimensions.

H1a: Economic dimension expectation: A pro/anti-EU dimension taps into economic and European integration issues.

H1b: Cultural dimension expectation: A left-right dimension taps into cultural issues.

Case selection and description

The goal of this article is to test the extent to which the recent interventions by the EU in the economies of national states led to a reconfiguration of the national political space. To this end we select an extreme case in terms of the independent variable. Hence, the EU intervention was far greater than in any other EU member state. Choosing an extreme case can also serve for exploratory analysis (Seawright and Gerring, 2008: 302). While the results of such analysis cannot be generalised to a larger population, it may serve as a template for further research.

Table 1 lists the five countries that have received bail-outs from the different European bail-out funds set up during the crisis. The total size of the bail-out compared to countries’ GDP differs strongly: Spain received a loan for less than a tenth of its GDP, Ireland and Portugal received loans for around a quarter of their GDP, Cyprus for half of its GDP and if one combines the two loans Greece received before the 2012 elections, these loans represent more than 100% of Greece’s GDP. Moreover, out of these five countries Greece was the only one that was bailed out more than once by the EU.¹ Without a doubt, Greece represents an extreme case of Europeanisation compared to other Eurozone members and even compared to other countries that were bailed out.

Table 1.

Impact of the bail-out per country.

Country	Date of bail-out	Percentage of GDP
Greece	May 2010	32
Greece	May 2012	70
Ireland	November 2010	24
Portugal	May 2011	29
Spain	July 2012	9
Cyprus	May 2013	47

Sources: Eurostat (2015), European Commission (2015), European Financial Stability Facility (2013), European Stability Mechanism (2012) and Eurogroup (2013).

Between 2009 and 2012 the nature of the Greek party system changed drastically. Before the crisis, there were two major parties in Greece, the left-wing PASOK and the right-wing ND. Additionally there were smaller parties of the left and the right: the communist KKE and the leftist SYRIZA and the right-wing populist LAOS. The parties and their relative sizes before the 2009 crisis are listed in Table 2. The key division in Greek politics was between the left and the right.

Table 2.

Political parties in Greece 2009–2014.

Name			Vote
Abb.	English	Greek	2009 (%)	2012 (May) (%)	2012 (June) (%)	Ideology
PASOK	Panhellenic Socialist Movement	Πανɛλλήνιο Σοσιαλιστικό Κίνημα	44	13	12	Social-Democracy
ND	New Democracy	Νέα Δημοκρατία	33	19	30	Conservatism
KKE	Communist Party of Greece	Κομμουνιστικό Κόμμα Ελλάδα	8	9	5	Communism
LAOS	Popular Orthodox Rally	Λαϊκός Ορθόδοξος Συναγɛρμός	6	3	2	Right-wing Populism
SYRIZA	Coalition of the Radical Left	Συνασπισμός Ριζοσπαστικής Αριστɛράς	5	18	27	Socialism
ANEL	Independent Greeks	Ανɛξάρτητοι Έλληνɛς	–	11	8	Right-wing Populism
XA	Golden Dawn	Χρυσή Αυγή	0	7	7	Radical nationalism
DIMAR	Democratic Left	Δημοκρατική Αριστɛρά	–	6	6	Social-Democracy

In Greece the meaning of left-right has been particularly flexible while it has consistently been the strongest predictor of party choice (Dinas, 2008; Freire and Costa Lobo, 2005; Karyotis and Rüdig, 2015). The convergence on the economic dimension of the two main parties, PASOK on the centre-left and ND on the centre-right, opened the door for new issues to enter the political agenda (Vasilopoulos and Vernardakis, 2011). At the time of the crisis issues such as immigration and nationalism had been established as part of the left-right dimension (Ellinas, 2013).

Despite this apparently simple division between left and right in the voter space that is represented by the struggle for power of the two main parties, ND and PASOK, Greek politics had a complex underlying dimensional structure in the party space even before the crisis. Looking at expert positioning of political parties, Bakker et al. (2012) found three dimensions structuring party positions: economic left-right, cultural left-right and EU integration. Gemenis and Dinas (2010) identified the same dimensions using manifesto data. The economic left-right is correlated with the EU dimension, while the cultural left-right is independent from both other dimensions (Bakker et al., 2012: 229).

The economic crisis had a strong effect on Greece: it led to the financial markets losing their faith in Greece’s ability to fulfil its financial obligations. In order to prevent bankruptcy, Greece turned towards the European institutions. In 2010 the country was given a bail-out loan accompanied by a memorandum that forced Greece to implement austerity measures, structural reforms and privatisation. The financial crisis became the catalyst that uncovered the weakness of the single currency: ‘it was a supranational agreement that had tied its members to a regime of mutual dependence and interconnectedness’ (Panagiotarea, 2013: 125). Of all the countries in the Eurozone, Greece has proven to be the most vulnerable to external shocks and required more than one bail-out package. What truly made Greece a special case was the weakness of its political institutions and the reluctance of politicians to change their policies (Afonso et al., 2014; Panagiotarea, 2013). The first bail-out was not enough; a second bail-out was necessary in 2012. While this second bail-out was being negotiated, the ruling PASOK government was replaced by a caretaker coalition of national unity (PASOK, ND and LAOS) led by a political independent. New elections were held in May 2012 and were narrowly won by ND. The results are shown in Table 2. After unsuccessful government formation talks, new elections were held in June 2012.

The economic crisis was the most salient issue in Greek politics with voters recognising the economy and unemployment as the two most important problems (Teperoglou and Tsatsanis, 2014). Support or opposition to the bail-out accordingly became the most prominent question in the 2012 elections (Dinas and Rori, 2013) as the parliamentary debates on the topic polarised both, the parties and voters. Economic issues no longer were valance issues and became positional issues as most voters could not trust any party to guide the country out of the crisis (Nezi and Katsanidou, 2014). All debates on economic policy are structured around the question whether Greece would honour its agreement and remain in the Eurozone or whether it would set its own financial policies and exit (Gemenis and Nezi, 2014).

Support for the bail-out among parties did not follow the traditional left-right dimension. A new division was created (Gemenis and Nezi, 2014): PASOK and the populist party LAOS supported the bail-out, while ND and a new left-wing party DIMAR took a more neutral position supporting the pro-bail-out government without openly positioning themselves in favour of the bail-out agreement. Opposition to the bail-out came from parties on the left and the right: the left-wing parties SYRIZA and KKE, the new right-wing populist party ANEL and the extreme right XA. The already existing connection between economic issues and EU issues may have intensified around the time of the 2012 election (Bakker et al., 2012: 229): the pro-European parties supported economic reforms, while the anti-European parties opposed austerity measures proposed by the EU.

Voting in the 2012 elections was determined by the bail-out division, and voters used their own position on the bail-out as a heuristic to navigate the crisis issues (Katsanidou, 2013; Teperoglou and Tsatsanis, 2014). In the public debate, the bail-out agreement had two components. On the one hand, it refers to the content of the solution to the crisis: the austerity, reform and privatisation policies put forward by the Troika. On the other hand, there is the discourse around the bail-out agreement touching upon national pride and continued membership in the EU and the Eurozone (Vasilopoulou et al., 2014). Political discourse portrayed the bail-out agreement as a loss of sovereignty and capitulation to external actors, and the blame was passed to the Greek government who was selling Greece out to those actors (Vasilopoulou et al., 2014). Eurosceptic parties and voters wanted Greece to leave the EU and return to the Drachma in order to regain the lost national pride.

At the same time, cultural issues were also high up on the political agenda; immigration in particular became a prominent issue (Teperoglou and Tsatsanis, 2014). Some voters perceived a new wave of immigrants from Middle East and Northern Africa as a threat to Greek national identity (Triantafyllidou and Kouki, 2013: 7). Protecting the ‘Greek way’ of life became an issue as the Greek government was unable to control its own borders (Dinas et al., 2013). XA, ANEL and LAOS were all positioned against immigration, while the ‘anti-immigrant’ agenda went mainstream as the two major parties, ND and to some extent PASOK, adopted it during the 2012 election campaign (Dinas and Rori, 2013). SYRIZA, DIMAR and KKE advocated a multicultural approach to immigration. Other issues also followed this left-right pattern such as law and order and civil liberties.

Methodology

Data sources and selection

Voting advice applications (VAA) offer a wealth of data (Wheatley, 2014: 5): for this study it is particularly useful because we need the placements of parties and voters on a large number of politically salient issues. It also allows us to use the exact same items for parties and for voters. We use the VAA Choose 4 Greece (C4G).

The party data consist out of party placement on 30 issues (Gemenis and Triga, 2012). Parties are positioned on basis of ‘judgemental’ coding of political text (e.g. party manifestos). In order to minimize random error and bias, the same issues were coded multiple times by expert coders (Gemenis, 2014). This method combines the best of expert judgement and textual analysis. The difference to the existing expert placement approach is that it allows various experts to anonymously ‘convince’ the other experts of their opinion by citing extracts from party documents.

The voter data come from the same VAA. It is crucial to note, however, that VAAs do not offer a representative sample of the population. One cannot draw inferences about the general population on the basis of data which are collected online and where users voluntarily opted to participate. Like most online surveys, VAA samples suffer from undercoverage and self-selection that could potentially bias estimates (Bethlehem, 2010). Research has shown that VAA users tend to be young, highly educated, male and more urban (Hooghe and Teepe, 2007; Wall et al., 2009).

In order to address the undercoverage of the selection, we construct a representative sample using sample matching procedure. Applying it to the data set of VAA users, we matched non-probability and probability samples on a respondent-to-respondent basis. (Daniel, 2012: 91). This method is used often to construct a sample that is equivalent to a representative sample from a large, but unrepresentative panel (Rivers, 2006). We apply this approach on a representative sample of Greek voters and match every respondent in this sample to a user from the VAA that is similar on all variables included in the VAA and the representative sample.

The voter data consist of 30 questions on policy issues as well as demographic data (gender, age, education level and region) and includes questions on past voting behaviour (in 2009) and intended voting behaviour in the 2012 elections.² We use the data set for the May 2012 elections that contains 75,249 entries.³ Fifty-nine per cent of these respondents answered both the substantive and the demographic questions completely. We can employ these to draw our matching sample. We work with the Greek sample from the Comparative Study of Electoral Systems (CSES, 2015) as a model for the representative sample, which has 1029 respondents. For each respondent in the representative sample of the CSES, we selected a user in the C4G database that had the same gender, education level⁴ and voting behaviour in the 2009 elections. For the region variable, we selected citizens from a region that had an identical level of unemployment in order to capture the centre–periphery divide and the specific economic environment in which the respondents find themselves. If there was more than one match, we selected the respondent that had the closest age. If there were multiple respondents with the same age, we selected the respondent randomly. Ten per cent of the respondents in the CSES could not be matched because no similar respondent could be found. These cases were left out of the analysis.

We use a matched data set and not a weighted data set because Mokken scaling can only be performed on the matched data: the existing Mokken package cannot deal with weights. Given that a weighted data set is considerably larger than the matched data set, results from the weighted data set are much more likely to be significant. The advantages of weighting over matching is that the results are more stable, in particular when an analysis concerns a variable with a high number of possible outcomes and a relatively small number of matched respondents. In the online appendix, we replicate the results of the regression analysis using the weighted sample previously used by Nezi and Katsanidou (2014).

Methods of data reduction

There are different ways to assess the dimensionality of a political space (De Vries and Marks, 2012). It is important to note, however, that the choice of scaling method has strong implications for the results that one gets. One therefore cannot claim that data has a ‘true’ dimensionality or that there is a ‘true’ spatial model that follows necessarily from the data (Benoit and Laver, 2012). In the end, data reduction is a process of creation (Coombs, 1964).

Following Otjes and Louwerse (2014) who also look at the dimensionality of answers in VAAs, this article uses Mokken scaling. This is a method from the item-response theory family. It tests whether respondents’ positions on a set of items form a single dimension (Mokken, 1971). The approach was specifically developed for dichotomous items, such true–false questions in an exam: Mokken scaling seeks to build a structure where items to which many respondents give the correct answer (‘easy items’) are on one side of the scale and items to which many respondents give the incorrect answer (‘difficult items’) are on the other side of the scale. If there is a single underlying structure that ranges from correct to incorrect, or from left to right, the items will form a consistent dimension. In order to determine whether this is the case, the H-value is calculated. This coefficient equals one when a set of items forms a perfect Mokken scale and zero when there is no statistical association between users’ answers on the items. Mokken (1971) suggests that scales with an H-value below 0.3 are too poor to be used. One can calculate the homogeneity coefficient for a scale as a whole (H) but also for individual items (H_i) and for pairs of items (H_ij).

Because our items are not dichotomous, we use polytomous Mokken scaling that can be applied to items with more than two response categories (Van der Ark, 2007). Note also that we use Mokken scaling in an exploratory way: we try to find strong dimensions based on parties’ and voters’ answers on the statements. The automated item selection procedure works as follows (Mokken, 1971): first, one selects the two items with the highest H_ij-coefficient; this is the starting pair for the first dimension. Then, one adds the item that results in the highest H-coefficient when added to the starting pair. This step is repeated until no items remain or the H-coefficient would drop below 0.3. If more than one item remains, the procedure will try to create a second dimension using the same procedure.⁵

We use the adaptation of the Automatic Item Selection Procedure (AISP) developed by Otjes and Louwerse (2014) that is neutral regarding the direction of the items. It is applied most often in survey research with large samples (but see Otjes and Louwerse, 2014). We now apply it to a sample of parties (N = 14). Straat et al. (2014) caution against the use of this method with small samples because of subpar accuracy. The results of the AISP should therefore be treated with considerable caution. In order to check the stability of the scaling results for the small samples of parties, we report the standard errors from the ‘Mokken’ package in R, which are based on the sample size. Moreover, we calculated standard errors based on a bootstrapping procedure, constructing 100 random samples from the items for which the H-value was calculated, and calculate the H-value for those samples. The mean H-value and the standard error of the H-value of these 100 samples are reported for political parties as well.

Methods of data analysis

In addition to testing the strength of our constructed dimensions, this article will also illustrate the empirical benefit of the chosen spatial model. To this end we will look at how voters’ positions on different dimensions relate to party preferences in the 2012 election. We use multinomial logistic regression to determine how positions on the different dimensions cohere with party preferences. The dependent variable is party preference before the election.⁶ The online appendix provides some descriptive variables for the C4G-matched sample. One should note that because these are party preferences asked during the 2012 campaign, a large share of voters is still undecided: therefore, we identify core supporters of these parties and not the swing voters. These are all in the ‘other’ category, which is also the base category. We use voters in this ‘other’ category as a base category in the regression because undecided voters are likely to be centrist, meaning that the coefficients reflect the ideological position of different voter groups vis-à-vis the centre. In order to aid interpretation of the regressions, all ratio/interval variables are normalised between zero and one, and we have plotted the results in figures using the ‘predict’ function from R.

Results

Scaling results. Table 3 shows the results of the AISP, the inductive scaling method for the VAA items, for both the voters and the parties. At the voter level, the first dimension includes more than half of all items (18). The second dimension taps into only five items.⁷ The first dimension mainly concerns the bail-out: whether citizens want to reform the economy to remain part of the EU or whether citizens oppose the memoranda, their economic implications and therefore membership of the EU itself. We can identify four categories within this dimension: four memoranda-related items are conceptually central. Voters either (a) agreed that the memoranda were necessary or (b) that they were equivalent to selling out the country. A second set of items concerns economic management: this includes all policy items in the memorandum, such as privatisation and pay cuts. But it also includes egalitarianism and co-funding of universities and university fees that were promoted by the PASOK government. A third set of items concerns EU integration, namely Greece’s Euro membership and the EU. There are two items that concern attitudes towards the Greek political system under pressure, the appointment of technocrats and the issue of demonstrations. Two final items show the capacity of the new dimension to integrate non-economic items. The second dimension concerns five items from the cluster immigration–integration–security–defence. These tap into a cultural understanding of the left-right dimension.

Table 3.

Scaling of voter and party positions.

Level	Voters		Parties
Scale	S1	S2	S1	S2	Memorandum
Inegalitarianism	0.37 (0.02)	–	0.81 (0.06)	–	No
Corporate taxes	–	–	0.86 (0.05)	–	No
Environmentalism	–	–	0.64 (0.16)	–	No
Solar energy	–	–	0.57 (0.14)	–	No
Privatisation	0.41 (0.02)	–	0.81 (0.07)	–	Yes
Police	–	0.41 (0.02)	0.81 (0.08)	–	No
Camera	0.33 (0.02)	–	0.79 (0.09)	–	No
Demonstrations	0.41 (0.02)	–	0.81 (0.07)	–	No
Citizenship	–	0.34 (0.03)	0.32 (0.20)	–	No
Multiculturalism	–	0.44 (0.02)	–	0.95 (0.02)	No
Flexible work	0.41 (0.02)	–	0.81 (0.06)	–	Yes
Co-funding universities	0.36 (0.02)	–	0.87 (0.06)	–	No
University fee	0.34 (0.02)	–	0.73 (0.10)	–	No
Hospitals	0.31 (0.02)	–	0.79 (0.06)	–	No
Civil servants	–	–	0.78 (0.08)	–	Yes
Pay cuts	0.37 (0.02)	–	0.75 (0.06)	–	Yes
Memoranda: crisis	0.43 (0.02)	–	0.80 (0.06)	–	Yes
Loans	0.47 (0.02)	–	0.81 (0.06)	–	Yes
Memoranda: selling	0.46 (0.02)	–	0.72 (0.10)	–	Yes
Memoranda: bankruptcy	0.50 (0.02)	–	0.84 (0.05)	–	Yes
Fiscal treaty	–	–	0.79 (0.06)	–	No
Grexit	0.36 (0.02)	–	0.84 (0.10)	–	Yes^a
Soft drugs	–	0.39 (0.02)	–	0.85 (0.08)	No
Defence	–	0.38 (0.03)	–	0.96 (0.02)	No
Professions	0.32 (0.02)	–	0.70 (0.12)	–	Yes
Church	–	–	0.34 (0.18)	–	No
Technocrats	0.38 (0.02)	–	0.75 (0.09)	–	Yes^b
Israel	0.32 (0.02)	–	0.77 (0.09)	–	No
FYROM	–	–	–	0.94 (0.05)	No
EU	0.41 (0.20)	–	0.71 (0.14)	–	No
H-value	0.39 (0.02)	0.39 (0.02)	0.74 (0.07)	0.92 (0.04)
H-value (bootstrapping)	–	–	0.74 (0.01)	0.94 (0.00)
Correlation with left-right placement^c	−0.44	0.57	−0.76	0.78
Correlation with ‘EU’^c,d	−0.76	−0.02	−0.73	−0.07
Mean	0.51	0.55	0.54	0.35
Median	0.51	0.54	0.56	0.13
Standard Deviation	0.19	0.21	0.31	0.40
Min.	0.03	0.00	0.11	0.00
Max.	1.00	1.00	1.00	1.00

Note: Unless indicated otherwise the values under S1 and S2 are H_ij values.

Grexit is a logical consequence in the case of non-compliance to the memorandum.

A technocratic government was not directly demanded in the memorandum but it was judged necessary for its implementation by the EU.

Correlation based on weighted data for voter level.

Specifically a bail-out dimension for voters and a pro/anti-EU dimension for parties.

The remaining items that were not attributed to any of the dimensions mainly concern economic policy, the public sector, foreign policy and the role of the church. These are items that one would generally expect to relate to some kind of general left-right dimension. This dimension, however, appears to be absent, as the only emerging dimensions from Mokken scaling relate either to (a) the national economy, EU membership and the role of international actors in the economy or (b) to cultural issues. The first dimension correlates strongly with the self-placement of users on the bail-out question. The second dimension shows a moderate correlation with left-right self-placement. The H-values for the entire dimensions are significantly higher than 0.3. The respondents in the matched sample conform to our expectations: their positions can be understood in terms of a bail-out dimension that taps into attitudes about the memoranda, the mandated reforms and economic governance, as well as a cultural left-right dimension that taps into attitudes on immigration and security.⁸

Some caution is warranted in interpreting the results at the party level, as the small number of cases can lead to unstable results. The AISP produces two dimensions. The first dimension includes 25 items tapping into a broad range of issues: the memoranda, economic management and EU membership, but also the environment and education. The second dimension includes five items, all concern cultural issues, immigration, law and order and foreign policy. Three of these five items overlap with the cultural dimension from the voter level. The H-values for both dimensions are far (and significantly) above the 0.3-threshold (with both methods). Two cultural items (concerning churches and citizenship) barely meet the 0.3 threshold and have quite high standard errors. We can correlate the party positions on the dimensions to the Chapel Hill Expert Survey (Bakker et al., 2015). The second dimension shows strong correlations with the general left-right item; the first dimension correlates equally with their pro-/anti-EU and left/right items.

Figure 1 shows the party positions on this two-dimensional space: parties are divided in four clusters: in the upper left corner we find the left-wing parties that opposed the bail-out: these are SYRIZA, DIMAR and the KKE. In the upper right corner, we find the right-wing anti-EU parties: XA, ANEL and LAOS. A third cluster is formed by ND, the right-wing pro-EU party. The PASOK has a left-wing pro-European position.

Figure 1.

Positions of the Greek political parties in a two-dimensional space.

At both the voter and the party level, one dimension concerns the economy and the memorandum. It includes 18 items for the voters and seven more for the parties. At the voter level, this is clearly a pro-/anti-bail-out dimension. At the party level, it taps into both the left-right and the pro-/anti-EU dimension. Both at the voter and the party level there is a separate cultural dimension that taps into the left/right divide. Both parties and voters support a similar interpretation of the political space: on the one hand, there is a dimension that taps into attitudes about European and economic matters and in particular international interference in Greece’s economic management. On the other hand, there is a dimension that taps into new cultural issues. All in all, the data corroborate our expectations.

Regression results

The dimensional structure resulting from our scaling exercise can be seen as credible if it demonstrates its value for explaining voter behaviour. Table 4 presents the result of a multinomial logistic regression analysis of vote intention for the seven parties that won seats in the 2012 elections. The reference category is voters that did not express a preference for these seven parties during the election campaign. Both our dimensions are significantly related to voter preference to most parties. The best way to understand those relationships is through visualisation: Figure 2 provides expected values of likelihood of preferring a particular party and voter position on both dimensions from the multinomial logistic regressions. The final entry provides a legend.

Table 4.

Multinominal logistic regression with matched data set without control variables.

Variable^a	ANEL	DIMAR	KKE	ND	PASOK	SYRIZA	XA
intercept	−6.35*** (0.80)	−1.12* (0.66)	−7.66*** (1.11)	−2.05*** (0.62)	1.30** (0.54)	−4.13*** (0.61)	−8.42*** (1.34)
Economic dimension	4.34*** (0.97)	0.32 (0.94)	7.63*** (1.29)	−5.89*** (1.01)	−5.95*** (0.99)	5.80*** (0.77)	3.19** (1.53)
Cultural dimension	2.99*** (0.78)	−2.91*** (0.86)	0.73 (0.92)	3.90*** (0.81)	−2.01** (0.84)	−1.51** (0.59)	5.47*** (1.36)

***

< 0.01 < ** < 0.05 < * < 0.01.

AIC = 2067, N = 927.

Reference category: other party preferences.

Matched data set.

Figure 2.

Estimated likelihood of voting for Greek parties. (a) PASOK, (b) ANEL, (c) DIMAR, (d) KKE, (e) ND, (f) SYRIZA, (g) XA, (h) legend.

The preference of voters follows the model we developed here: core supporters for the PASOK and ND have similar views on the economic dimension: voters of these two parties support Greece’s membership in the EU and the implementation of the memoranda. The supporters of these two parties differ on their views on cultural issues: PASOK has progressive supporters and the New Democracy has conservative supporters. On the other side of the economic dimension we find the supporters of the KKE, SYRIZA, XA and ANEL. The supporters of these parties oppose the memoranda and their economic consequences and are against EU membership. The supporters of ANEL, XA and SYRIZA differ in their cultural views: supporters of ANEL and XA are more conservative and SYRIZA’s supporters are more progressive. The XA supporters are primarily defined by their cultural views and less by their economic views. The KKE gets the support of voters who oppose the bail-out regardless of their attitudes towards cultural issues. The supporters of DIMAR are not defined by their views on economic issues: its supporters have progressive views on law, order and immigration. As expected, the pro-EU parties attracted pro-EU voters while Eurosceptic parties attracted Eurosceptic voters. The culturally right-wing parties attracted culturally right-wing voters. The online appendix shows that the same substantial conclusions can be drawn when adding control variables and when examining a weighted sample.

Conclusion

This article sketched the transformation of the Greek party system under the pressure of the European sovereign debt crisis. It analysed a transformational moment in Greek politics, but also a transformational moment in European politics. We found that, unlike what previous studies had proposed (Kriesi et al., 2008; Mair, 2000), European integration can restructure the way parties and voters position themselves on economic issues. On basis of Mokken scaling procedures for the 30 items that were included in the VAA Choose 4 Greece, we constructed two-dimensional models of the Greek voter and party space. These models were shown to be internally cohesive and were externally validated by illustrating their value for voting behaviour. The two dimensions structuring this space were: (a) an economic dimension that also concerned European integration, the bail-out memoranda and positions on the mandated economic reforms. There was a significant correlation between pro-/anti-EU and economic positions for the parties and the pro-/anti-bail-out self-placement for the voters; and (b) a cultural left-right dimension that centred on cultural issues from the cluster immigration–integration–security–defence. At the party level the first dimension also tapped into other economic, environmental and ethical issues, and it showed a strong correlation with expert placement of parties on the left-right dimension. Both the cultural and the economic dimensions predict voting behaviour for all major parties: the bail-out dimension is a stronger predictor of voter choice than the cultural dimension.

The relevance of this study goes beyond the borders of the Greek case and increases as European integration deepens. This is true not only because one might expect similar patterns of response to the Eurozone crisis in other countries that signed bail-out agreements, but because the crisis made clear that EU membership came attached with financial obligations. In order to ensure the stability and longevity of the Euro, the EC has expanded its competences over budgetary policy. EU and Eurozone membership constrains policy choices of European governments. The focus of the Stability and Growth Pact on fiscal responsibility limits the ability of parties in Eurozone countries to offer competing alternatives for budgetary policy without raising the issue of Eurozone membership. Indications are that in a country that is much less affected by the economic crisis, namely the Netherlands, a similar pro-/anti-EU dimension in part structures party policy positions (Otjes, 2015). Future research could pursue the question whether the patterns found here can be found in other Eurozone countries, in particular those also strongly affected by the crisis. The rise of parties critical to austerity and the role of the EC such Podemos in Spain and the Five Star Movement in Italy indicates that this is very likely.

Other researchers may also want to examine whether the new political space persists in the Greek party system over time. Early indications are that the new dimensions continue to play a role: the January 2015 electoral campaign of SYRIZA was based on the, until then, unimaginable combination of rejecting austerity and staying in the Eurozone, which got SYRIZA in government. SYRIZA formed a coalition with the Eurosceptic, anti-austerity but culturally right-wing Independent Greeks after the 2015 Greek election. The newly elected government based their negotiations on the argument that austerity did not work and that Europe should change from within. This argumentation failed to find supporters in the Eurogroup. With the referendum on 5 July 2015 they put their argument to the voters: 61% of the voters rejected the deal offered by the Eurogroup; the government assured that rejecting the deal would not lead to Greece’s exit from the euro or the EU. The government soon thereafter accepted a harsher memorandum between Greece and the European institutions, which was backed by the vast majority of the Greek parliament including most government (and until then anti-austerity) MPs. This shows that even anti-austerity parties with the priority of staying in the Eurozone cannot ignore the budgetary policy set by the EC. It is the final end of the belief that ‘the economy’ and ‘Europe’ can be considered separately from each other.

Footnotes

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Afonso

Zartaloudis

Papadopoulos

(2014) How party linkages shape austerity politics: Clientelism and fiscal adjustment in Greece and Portugal during the eurozone crisis. Journal of European Public Policy 22(3): 315–334.

Bakker R, De Vries C, Hooghe L, et al. (2015) Measuring party positions in europe: The chapel hill expert survey trend file, 1999–2010. Party Politics 21(1): 143–152.

Bakker

Jolly

Polk

(2012) Complexity in the European party space: Exploring dimensionality with experts. European Union Politics 13(2): 219–245.

Benoit

Laver

(2012) The dimensionality of political space: Epistemological and methodological considerations. European Union Politics 13(2): 194–218.

Bethlehem

(2010) Selection bias in web surveys. International Statistical Review 78(2): 161–188.

Coombs

(1964) A Theory of Data, Hoboken, New York: Wiley.

The Comparative Study of Electoral Systems (2015) CSES Module 4: 2011–2016. DOI: 10.7804/cses.module4.2015-03-20. Available at: http://www.cses.org/datacenter/module4/module4.htm (accessed 7 November 2014).

Daniel

(2012) Sampling Essentials. Practical Guidelines for Making Sampling Choices, Thousand Oaks: Sage.

De Vries

Marks

(2012) The struggle over dimensionality: A note on theory and empirics. European Union Politics 13(2): 185–193.

10.

Dinas

(2008) Big expectations, small outcomes: The impact of leaders’ personal appeal in the 2004 Greek election. Electoral Studies 27(3): 505–517.

11.

Dinas E, Georgiadou V, Konstantinidis I, et al. (2013) From dusk to dawn. Local party organization and party success of right-wing extremism. Party Politics. Epub ahead of print 1 December 2013. DOI: 10.1177/1354068813511381.

12.

Dinas

Rori

(2013) The 2012 Greek parliamentary elections: Fear and loathing in the polls. West European Politics 36(1): 270–282.

13.

Ellinas

(2013) The rise of the Golden Dawn: The new face of the far right in Greece. South European Society and Politics 18(4): 543–565.

14.

Eurogroup (2013) Eurogroup statement on Greece. Available at: www.consilium.europa.eu/uedocs/cms_Data/docs/pressdata/en/ecofin/136767.pdf (accessed 5 May 2015).

15.

European Commission (2015) ‘Financial assistance to Greece’. Available at: http://ec.europa.eu/economy_finance/assistance_eu_ms/greek_loan_facility/index_en.htm (accessed 5 May 2015).

16.

European Financial Stability Facility (2013) Frequently asked questions European Financial Stability Facility. Available at: www.efsf.europa.eu/attachments/faq_en.pdf (accessed 5 May 2015).

17.

European Stability Mechanism (2012) Frequently asked questions Spain. Available at: www.esm.europa.eu/pdf/FAQ%20spain%2007120212.pdf (accessed 5 May 2015).

18.

Eurostat (2015) GDP and main components. Available at: http://ec.europa.eu/eurostat/web/products-datasets/-/namq_10_gdp (accessed 5 May 2015).

19.

Evans

Heath

Lalljee

(1996) Measuring Left-Right and Libertarian-Authoritarian values in the British electorate. The British Journal of Sociology 47(1): 93–112.

20.

Feldman

Conover

(1983) Candidates, issues and voters: The role of inference in political perception. The Journal of Politics 45(4): 810–839.

21.

Freire

Costa Lobo

(2005) Economics, ideology and vote: Southern Europe, 1985–2000. European Journal of Political Research 44(4): 493–518.

22.

Fuchs

Klingemann

H-D

(1989) The left-right schema. In: Jennings

van Deth

Barnes

(eds) Continuities in Political Action: A Longitudinal Study of Political Orientations in Three Western Democracies Vol. 5, Berlin: Walter de Gruyter, pp. 203–234.

23.

Gabel

Hix

(2002) Defining the Eu political space: An empirical study of the European elections manifestos, 1979–1999. Comparative Political Studies 35(8): 934–964.

24.

Gemenis

(2014) An iterative expert survey approach for estimating parties’ policy positions. Quality and Quantity 49(6): 1–16.

25.

Gemenis

Dinas

(2010) Confrontation still? Examining parties policy positions in Greece. Comparative European Politics 8(2): 179–201.

26.

Gemenis

Nezi

(2014) Government-opposition dynamics during the economic crisis in Greece. Journal of Legislative Studies 21(1): 14–34.

27.

Gemenis K and Triga V (2012) Party positions for the May and June 2012 parliamentary elections in Greece. DANS. Available at: https://dx-doi-org.web.bisu.edu.cn/10.17026/dans-xtg-5gup (accessed 15 July 2015).

28.

Gerring

(2007) Case Study Research. Principles and Practices, Cambridge: Cambridge University Press.

29.

Hinich

Munger

(1996) Ideology and the Theory of Political Choice, Ann Arbor, MI: University of Michigan Press.

30.

Hix

Lord

(1997) Political Parties in the European Union, Basingstoke: Palgrave Macmillan.

31.

Hooghe

Marks

Wilson

(2002) Does Left/Right structure party positions on European integration? Comparative Political Studies 35(8): 965–989.

32.

Hooghe

Teepe

(2007) Party profiles on the web: An analysis of the logfiles of non-partisan interactive political internet sites in the 2003 and 2004 election campaigns in Belgium. New Media and Society 9(6): 965–985.

33.

Inglehart

Klingemann

H-D

(1976) Party identification, ideological preference and the Left–Right dimension among Western mass publics. In: Budge

Crewe

Farlie

(eds) Party Identification and Beyond: Representations of Voting and Party Competition, London: John Wiley, pp. 243–276.

34.

Karyotis

Rüdig

(2015) Blame and punishment? The electoral politics of extreme austerity in Greece. Political Studies 63(1): 2–24.

35.

Katsanidou

(2013) The Euro crisis and new dimensions of contestation in national politics. In: De Witte

Heritier

Trechsel

(eds) The Euro Crisis and the State of European Democracy, Florence: European University Institute, pp. 242–263.

36.

Kriesi

(2014) The populist challenge. West European Politics 37(2): 361–378.

37.

Kriesi

Grande

(2014) The Europeanization of the national political debate. In: Cramme

Hobolt

(eds) Democratic Politics in a European Union Under Stress, Oxford: Oxford University Press, pp. 67–86.

38.

Kriesi

Grande

Dolezal

(2012) Political Conflict in Western Europe, Cambridge: Cambridge University Press.

39.

Kriesi

Grande

Lachat

(2008) West European Politics in the Age of Globalization, Cambridge: Cambridge University Press.

40.

Ladrech

(2002) Europeanization and political parties towards a framework for analysis. Party Politics 8(4): 389–403.

41.

Laponce

(1970) Note on the use of the left-right dimension. Comparative Political Studies 2(4): 481–502.

42.

Lefkofridi

Wagner

Willmann

(2014) Left-Authoritarians and policy representation in Western Europe: Electoral choice across ideological dimensions. West European Politics 37(1): 65–90.

43.

Mair

(2000) The limited impact of Europe in national party systems. West European Politics 23(4): 27–51.

44.

Mair

(2008) Political parties and party systems. In: Graziano

Vink

(eds) Europeanization. New Research Agendas, Basingstoke: Palgrave Macmillan.

45.

Mokken

(1971) A Theory and Procedure of Scale Analysis, Berlin: De Gruyter.

46.

Nanou

Dorussen

(2013) European integration and electoral democracy: How the European Union constrains party competition in the Member States. European Journal of Political Research 52(1): 71–93.

47.

Nezi

Katsanidou

(2014) From valence to position: Economic voting in extraordinary conditions. Acta Politica 49(4): 413–430.

48.

Otjes S and Katsanidou (2015) How the European debt crisis reshaped national political space: Replication materials. Datorium. Available at: https://dx-doi-org.web.bisu.edu.cn/10.7802/1086 (accessed 18 August 2015).

49.

Otjes SP (2015) How the Eurozone crisis reshaped the national economic policy space: The Netherlands 2006–2012. Acta Politica. 19 June 2015. DOI: 10.1057/ap.2015.11.

50.

Otjes

Louwerse

(2014) Spatial models in voting advice applications. Electoral Studies 36(4): 263–271.

51.

Panagiotarea

(2013) Greece in the Euro: Economic Delinquency or System Failure?, Colchester: ECPR Press.

52.

Rivers D (2006) Sample matching. Representative sampling from Internet panels mimeo YouGovPolimetrix.

53.

Seawright

Gerring

(2008) Case selection techniques in case study research a menu of qualitative and quantitative options. Political Research Quarterly 61(2): 294–308.

54.

Straat

Van der Ark

Sijtsma

(2014) Minimum sample size requirements for Mokken scale analysis. Educational and Psychological Measurement 74(5): 809–822.

55.

Teperoglou

Tsatsanis

(2014) Dealignment, de-legitimation and the implosion of the two-party system in Greece: The earthquake election of 6 May 2012. Journal of Elections, Public Opinion and Parties 24(2): 222–242.

56.

Triantafyllidou

Kouki

(2013) Muslim immigrants and the Greek nation: The emergence of nationalist intolerance. Ethnicities 13(6): 709–728.

57.

Van der Ark

(2007) Mokken scale analysis in R. Journal of Statistical Software 20(11): 1–19.

58.

Van der Brug

Van Spanje

(2009) Immigration, Europe and the ‘new’ socio-cultural dimension. European Journal of Political Research 48(3): 309–334.

59.

Vasilopoulos P and Vernardakis C (2011) The rise and fall of the Greek Conservative Party: Ideological realignments and egocentric economic voting at the dawn of the financial crisis. In: Paper prepared for presentation at the 61st Political Studies Association Annual Conference, London, United Kingdom, 19–21 April 2011.

60.

Vasilopoulou

Halikiopoulou

Exadaktylos

(2014) Greece in crisis: Austerity, populism and the politics of blame. Journal of Common Market Studies 52(2): 388–402.

61.

Wall

Sudulich

Costello

(2009) Picking your party online–An investigation of Ireland’s first online voting advice application. Information Polity 14(3): 203–218.

62.

Wheatley J (2014) Identifying latent policy dimensions from public opinion data: An inductive approach. Journal of Elections, Public Opinion and Parties 25(2): 215–233.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

9.35 MB

0.00 MB

0.13 MB