Parties Acting Strategically: National Parliaments’ Role in Holding the EU Accountable

Abstract

The existing assessments of EU politicization have predominantly focused on the media and in the actors that are reported in that arena. This chapter offers a novel approach to the study of EU politicization by focusing, instead, on the institutional arena of six EU countries. Using the MAPLE dataset of parliamentary debates, with a total of 724,963 speeches covering a period of 20 years, this chapter offers a comparative and longitudinal analysis of EU politicization at the institutional level, assessing the impact that the Eurozone crisis had in the parliamentary debates of the EU. Automated and manual content analysis methods are used to answer three main questions: Has the EU been politicized at the institutional level? Did the Eurozone crisis affect the politicization of the EU in parliaments? What characteristics of political parties have determined the salience and contestation of the EU topic? Our results suggest that the EU has not been, in the past 20 years, a very salient nor contested topic at the institutional level. However, we find that the parties’ contribution to the democratic legitimacy of the European integration varies significantly, with the parties’ position on the EU topic being the strongest determinant of EU politicization.

Introduction

The existing literature assessing the magnitude of EU politicization has focused predominantly on the intermediary arenas of political communication. In line with some of the main findings of the previous chapter, those studies have shown us that EU politicization occurs on a punctuated basis in certain countries/regions, amid important EU-related events. Differently, however, the degree to what EU issues has been politicized in the institutional arena remains largely untapped in a comparative and longitudinal basis. This chapter addresses this lack of research by offering an extensive and comparative analysis of EU politicization in the parliamentary arenas of six EU member-states.

Using an original dataset of parliamentary plenary speeches from the six MAPLE countries, over a period of 20 years, this chapter explores the following three research questions: (1) To what degree has the EU been politicized, in terms of its salience and contestation, in the parliamentary debates of different EU countries? (2) Did the Eurozone crisis impact the frequency and the way in which the European integration dimension is discussed at the institutional level? (3) what party characteristics, namely, left–right ideology, incumbency and EU position, are associated with higher levels of EU politicization? We explore these questions by combining automated and manual methods of content analysis, applied to a corpus of 724,963 speeches.

Our research questions matter for at least three main reasons. The first one concerns, as we mentioned, the lack of longitudinal and cross-national research on how the EU has been debated in this arena, particularly covering the period following the onset of the Eurozone crisis. The existing studies on EU politicization not only have mainly focused on other channels but also, when looking at the institutional arena, the literature mainly deals with the parliament’s legislative and oversight roles, rather than its communication function. While the usefulness of mainstream media in understanding societal interests, and their levels of contestation, is indisputable, national parliaments remain the most relevant and consequential platform to debate complex political issues such as the regional integration process.

The second reason concerns the anticipated differences, in terms of political communication, between institutional and intermediary arenas. Even though the main political actors in media and parliaments are largely the same, the two arenas have different operating logics, and rules, that can distinctly shape the EU’s salience and contestation. This is a crucial aspect since the media logic influences the content of the news coverage of political events. Therefore, a key aspect of the parliaments’ communication function is offering an unmediated measurement of EU politicization and an undistorted picture of how the EU has been discussed by the citizens’ elected representatives. An analysis that, therefore, complements the existing studies of media content, offering us a more encompassing understanding of EU politicization.

Finally, the third reason concerns the increasing importance attributed to parliamentary debates for the democratic legitimacy of the European polity. When it comes to the different modes of representation in the EU, the national parliaments, and the deliberation that occurs in them, are the most important foundations for the emergence of a European ‘demoi-cracy’. In this sense, the comparative analysis made in this chapter is important not only to offer a more nuanced, or alternative, assessment of the magnitude of EU politicization in different European countries but also to corroborate envisioned expectations regarding the role of national parliaments in furthering the European integration process. To put it in other words, as it is argued in the introduction of this book, the focus on national political institutions is crucial to understand how the growth in Europeanization is being legitimized. The analysis in this chapter not only serves that purpose but is also a fundamental contribute to the main goal of this book, which is to understand the effect of EU politicization on national elections.

This chapter is structured into five sections. After this first introductory part, the second section of the chapter offers a brief literature review of EU politicization in national parliamentary debates, focusing on its magnitude and determinants. The third section of the chapter presents the research questions and expectations of this study, as well as the data and methods used to investigate them. After that, the main results of our analysis are presented and discussed in the fourth section of the chapter. Finally, in the last section, we summarize our main findings and discuss some of their potential implications.

Literature Review

National parliaments are, unquestionably, the cornerstones of representative democracy and political contention in the parliamentary systems of the different EU countries. Yet, the two main conceptions of the European integration process (i.e., intergovernmental and supranational) have always attributed a secondary/indirect role to national parliaments regarding their involvement in the EU policy-making (Wendler, 2016). These two main ‘schools of thought’, or lenses to study regional integration, have focused either on the role of national governments or the supranational institutions and elites of the EU (Schimmelfennig & Rittberger, 2006; Tsebelis & Garrett, 2001). Moreover, the perceived decline of parliaments’ capacity to carry out their functions in Western democracies (Döring, 1995a) might have contributed as well to the lack of attention given, by the literature, to the national parliaments’ involvement in the European integration process. However, this situation started to change in the early 2000s with the Laeken declaration and the respective commitment of the EU to a ‘greater democracy, transparency and efficacy’.¹

The increasing debates, and concerns, surrounding the lack of democratic legitimacy led to a distinct formulation of the European polity, as a demoi-cracy (Besson, 2006; Nicolaïdis, 2004), that placed considerably more importance in the relationship between national parliaments and the EU.² This became particularly evident with the Lisbon Treaty, which gave national parliaments a more important role in EU affairs by strengthening their scrutiny and participation rights (Auel et al., 2018; Nicolaïdis, 2013). The public debates are an essential aspect of this emerging conception of Europe, which demands, as well, a more important communicative role of parliaments. In this view, the national parliament became a central arena for public reflection and debate about the EU (Wendler, 2016), which, in a way, has sparked the interest of EU scholars on the, somehow overlooked, communication function of those institutions. To put it in other words, the communication and information role of parliaments has become a key dimension for the study of EU politicization.

While acknowledging the multifaceted and resourceful nature of the parliaments, this chapter focuses on the parliaments’ role as arenas of political communication. In that context, the public debates in the plenary are, notwithstanding the question of whether citizens actually follow them,³ the most important means to fulfil the information and communication functions of parliaments (Auel & Raunio, 2014a, 2014b). As Auel and Raunio (2014a, p. 4) put it, the plenary debates offer a concrete arena for both ‘articulating and representing societal interests and informing the electorate about issues on the political agenda’. These debates are therefore important for citizens to hold political parties accountable for their positions on the EU, even though the parties have some strategies available to them, such as ‘blame avoidance’ and ‘credit claiming’, that can still blur that accountability (Lord & Pollack, 2010; Hobolt & Tilley, 2014). To put it differently, the plenary debates are both an important and unique setting to access the magnitude of EU politicization since they offer, in the utmost consequential setting, an unmediated discussion of the most pressing societal issues, as well as the stance of the key political parties on them. For those reasons, most of the existing studies assessing the politicization of the EU at the institutional level have focused on plenary debates (e.g., Auel & Raunio, 2014b; Rauh, 2015; Rauh & de Wilde, 2018).

When it comes to the magnitude of EU politicization in national parliaments, we can find some opposing views in the literature. For some authors, on the one hand, it seems indisputable that European integration has always been politicized at the institutional arena, even though those high levels of contestation might not be reflected in the political actors’ communication addressed to the citizens (Hurrelmann et al., 2015, p. 46). Somewhat in line with this idea, Wendler (2016) found that normative EU claims are both more frequent and more contentious in parliamentary debates, compared to more pragmatic ones.

On the other hand, the strong incentives for parties to strategically depoliticize the European integration process are also very well known (see Mair, 2000) and, to a great extent, also more empirically substantiated. In fact, some studies have shown that most national parliaments do not ‘live up to their task of bringing “Europe” closer to the citizens’ (Auel & Raunio, 2014a, p. 10). Using an automated content analysis of parliamentary speeches in four EU countries, from 1991 to 2015, Rauh and de Wilde (2018) also showed that, despite being somehow responsive to supranational decisions and events, there is a lack of balanced debates on European integration and a limited supply of electoral choice. Looking at four countries (Finland, France, Germany, and UK), between 2002 and 2012, Auel and Raunio (2014b) find that, despite some cross-country variation, EU politicization has been generally low in parliamentary debates. However, these authors observe an increase in EU politicization in the last two years of their analysis (2011 and 2012), particularly in the Finnish Eduskunta.

Overall, the perceptions of high levels of EU politicization at the institutional level seem to be based more on conventional wisdom than thorough comparative research. As we saw, despite considerable variation between countries and parties (Lauwers et al., 2021), the existing longitudinal and cross-country research suggests that the levels of EU politicization in parliaments have been, overall, relatively low, particularly when it comes to the sheer volume of debates. Moreover, similar to what we see in the media (Hutter et al., 2016), the EU becomes more salient in parliamentary debates during periods around important EU events (Rauh & de Wilde, 2018). In this sense, in line with what research found for media (Hutter & Kriesi, 2019, Silva et al., 2022), the eurozone crisis might have contributed to higher levels of EU politicization also in the parliaments. Some studies have indeed pointed in this direction, suggesting that EU politicization increased in the parliaments’ debates (Auel & Raunio, 2014b; Riekmann & Wydra, 2013) and activities (Auel & Höing, 2014) with the eurozone crisis. Other authors, with more encompassing assessments, have suggested a less significant impact of the crisis in the parliaments’ politicization of the EU (e.g., Rauh & de Wilde, 2018). Overall, in the institutional arena, the impact of the eurozone crisis in the magnitude of EU politicization, across countries and political parties, remains a puzzling and inconclusive subject.

Regarding the comparative assessment of EU politicization in the institutional arena, and the possible impact that the eurozone crisis had on it, the literature has taken into consideration two important dimensions or categories of factors. The first one concerns the existence of considerable differences between parliaments, arguably much higher than in the media, that make cross-national comparisons not only important but far more challenging (Döring, 1995a). When it comes to the parliamentary debates, institutional factors such as the governments’ level of agenda control (Döring, 1995b), type of legislature, being a ‘debating’ or ‘working’ parliament (Dann, 2003) and degree of EU affairs delegated to committees (Auel & Raunio, 2014b), can potentially shape how much, and in which way, the EU is debated. The literature has found, however, a rather negligible impact of institutional factors on the levels of EU politicization (Auel & Raunio, 2014b). Instead, actors seem to be the most important factor in explaining the levels of EU politicization in parliaments.

The second important dimension in the assessment of EU politicization, at the institutional level, concerns the main political actors and the differences between parties in terms of their characteristics and respective roles in the legislature. More concretely, strategic considerations, and therefore government/opposition, left/right and pro-/anti-EU differences seem crucial for the (de)politicization of the EU issues in national parliaments (Kaniok & Brusenbauch Meislova, 2021; Lupato, 2014; Navarro & Brouard, 2014). Similarly, Wendler (2016) also found that EU contestation in parliamentary debates is mainly determined by the existence of Eurosceptic parties, being institutional arrangements, and public opinion on the EU, rather insignificant dimensions.

When it comes to the discussion of EU topics, some studies also found that incumbent parties outperform the challenging ones (Rauh & de Wilde, 2018). Moreover, a far more puzzling finding from Rauh and de Wilde (2018) is that opposition parties seem to debate the EU less when citizens’ Euroscepticism increases. In line with that, Auel and Raunio (2014b) found that, overall, the EU salience is higher in contexts where party conflict and public Euroscepticism are both low. In this sense, differently from what we saw for the media in the previous chapter, the correlation between EU salience and contestation seems to be negative in the parliaments, which contributes to a lack of EU accountability in this arena.

Finally, moving beyond the magnitude of politicization, the aspects of the European Integration discussed in the parliaments can also have important implications for the EU issue voting. According to the literature, there are different dimensions of the EU that can be the object of political contestation (Mair, 2007). In this regard, Hurrelmann et al. (2015) offer a useful typology that categorizes four types of EU debates. The first dimension, membership, concerns the creation, consolidation, and geographical reach of the EU institutions. The second one, constitutional structure, deals with debates surrounding the procedures and responsibilities of the different EU institutions, as well as the EU’s decision-making process in general. Finally, the last two dimensions deal both with the reach and penetration of EU legislation in the member-states. While one, EU policies, focus on policies emanated from EU institutions, the other, domesticated policies, concerns national-level policies that are somehow a consequence of the European integration process.

Overall, regarding the magnitude of EU politicization in parliamentary debates, the existing literature has echoed two key aspects. The first one is that existing differences between parliaments seem to be determined mainly by party-level characteristics rather than institutional factors or levels of EU politicization in the other two arenas (i.e., citizens and media). In fact, looking at three EU countries (Denmark, Ireland, and the Netherlands), de Wilde (2014) finds that, when it comes to the politicization of Europe, differences between arenas (i.e., media and parliaments) are more substantive than cross-national and diachronic variation. The second aspect is that the Eurozone crisis, differently from the media, might not have contributed to a higher politicization of the EU in the institutional arena. In this sense, if EU politicization did not increase, the crisis might have contributed to a deficit of democratic accountability in the EU member-states. These two aspects are explored in this chapter with a longitudinal analysis of EU politicization in the parliamentary debates of the six MAPLE countries.

Expectations, Data, and Methods

To answer our research questions, we develop and test five expectations, based on existing literature, regarding the results of our analysis of parliamentary debates. First of all, we expect that the magnitude of EU politicization in parliamentary debates will vary more across countries/parliaments than across time. The literature has consistently showed that the overall levels of EU politicization in parliaments can differ considerably between countries, being particularly higher in places where, intriguingly, the EU is a less contested issue. This is justified by the fact that it is less risky for parties to debate the EU when the electorate has a predominantly favourable view of the European Integration process. In this regard, concerning the MAPLE countries, literature has found the salience of the EU to be comparatively higher in the German parliament (Auel & Raunio, 2014a, 2014b, Rauh & de Wilde, 2018) and lower in the parliaments of Spain (Lupato, 2014) and Greece (Bélanger & Schimmelfennig, 2021). The situation in the remaining countries is much less known from a comparative perspective, making, therefore, this chapter’s analysis even more important.

The second expectation concerns the impact of the crisis. Similar to what was found for the media in the previous chapter, we also anticipate, in the parliaments, an increase in EU salience and contestation after the Eurozone crisis. As we mentioned, there are some hints of evidence in the literature to the fact that the EU politicization increased in parliamentary debates after that crisis (Auel & Raunio, 2014b; Riekmann & Wydra, 2013), even despite none of those studies includes one of the debtor countries (i.e., Greece, Ireland, Portugal, and Spain). However, at the same time, the salience of EU in parliaments also seems to be negatively associated with its contestation levels. These seemingly contradictory findings suggest, as we anticipate for this chapter, that the eurozone crisis had a more noticeable impact on the levels of contestation rather than on the salience of the EU dimension.

The third expectation concerns the dimensions of the EU that are more salient in the parliamentary debates. On the one hand, the multifaceted nature of parliaments and its diversity of actors should stimulate a higher variety of debates about the EU, which is also important for the legitimization of the European integration process. On the other hand, as Bartolini (2005, p. 349) discusses, because political parties lack the capacity to debate constitutive issues of the EU, they seem to have adopted a strategy of ‘depoliticization’ of that specific dimension (Mair, 2007). Therefore, similarly to what we observed for the media in the previous chapter of this book, we also anticipate that the EU debates, in the national parliaments, have remained predominantly in the realm of policies, instead of dealing with aspects related to EU membership and constitutive issues.

The fourth expectation concerns the anticipated differences, between types of parties, in the levels of EU politicization. More concretely, we explore three different distinctions. The first one is government versus opposition. In this regard, we anticipate that government parties mention the EU more frequently in their interventions in the parliament (Lupato, 2014; Rauh & de Wilde, 2018). The second is the position of the party towards the EU. Here we expect that Eurosceptic parties are likelier to politicize the EU in their parliamentary speeches (Navarro & Brouard, 2014; Wendler, 2016). Finally, the third dimension considered is the ideological leaning (left–right) of the parties. While we do not have a concrete expectation regarding this dimension, we do expect to observe distinct patterns across the six countries analysed regarding this distinction.

The fifth, and final, expectation concerns the relationship between salience and contestation. Some studies have found, at the country level, that the salience of EU is higher in countries that show lower levels of EU contestation (e.g., Auel & Raunio, 2014b). In line with that, we also expect that the countries with higher levels of EU salience are also the ones where the proportion of articles with a negative tone towards the EU is lower.

Our research question and expectations are examined using an original dataset of plenary sessions’ speeches from six different European parliaments: Chambre des Représentants (Belgium), Bundestag (Germany), Vouli (Greece), Oireachtas (Ireland), Assembleia da República (Portugal), and Congreso de los Diputados (Spain). The dataset includes all the speeches with more than 40 words⁴ from a period of 20 years (see Table 3.1). The dataset includes, for each speech, relevant information such as its date, speaker’s name, and the party that he or she belongs to. In total, the final dataset used for this analysis includes 724,963 unique speeches.

Table 3.1 Total number of speeches per country and period

Our assessment of EU politicization in parliaments focuses on two dimensions of the concept: salience and contestation.⁵ We operationalize EU salience as the percentage of speeches, in a particular year or legislative term, that mention the EU. We identify the EU speeches using an extensive list of EU-related terms (Appendix: Table 3.2). This list, translated into seven languages, was adapted from the codebook of Maier et al. (2014). Overall, about 12% (86,751) of the speeches in our dataset mentioned the EU.

The dimension of EU contestation was operationalized as the proportion of speeches in a year, or term, that had a negative tone towards the EU. The process to calculate this score was the following. The first step was extracting, for each speech, all sentences mentioning the EU. Since our dataset includes speeches in seven different languages, we then translated those EU sentences into a single ‘pivot’ language (Lucas et al., 2015), using the R package googleLanguageR(), which gives users access to Google’s translation Application Programming Interface (API) service. After that, the EU sentences’ tone was measured using the algorithm provided by the package sentimentr(),⁶ by Rinker (2019). A speech was considered negative if the average of all its EU sentences’ tone values/scores was lower than zero. Similar to the previous chapter, we assume that the contestation of EU increases, in a particular year or term, when the share of negative articles also increases.

The different manifestations of the EU discussed in the parliamentary speeches were manually coded by a team of trained graduate students that identified, from a random sample of speeches mentioning the EU, if the speech dealt with (1) membership; (2) constitutional structure, (3) EU policies, or (4) domesticated policies (Hurrelmann et al., 2015, p. 45).⁷ Additionally, each party with parliamentary representation was also coded, for the three categories of interest, using Chapel Hill expert surveys. These categories were whether the party was in government, the CHES general left/right variable, and, lastly, the parties’ stance towards the EU.

Finally, to better understand the determinants of EU politicization at the party level, and further test our main expectations, we have also conducted a series of linear regression analyses with our dataset. The two dependent variables used in the models were EU Salience and EU Contestation. The unit of analysis in the models was the party score for each plenary session and includes all political parties with parliamentary seats. In addition to our variables of interest (EU stance, Ideology, and Incumbency) we also include, as controls, the parties’ seat share, if the party is from a debtor country and if the plenary session occurred after the Eurozone crisis (2009). Additionally, the third and fourth model of each table also includes, respectively, country dummies and an election variable, controlling for the last year of each plenary session, to analyse a time trend. All variables included in the models were normalized to vary between 0 and 1.

Results and Discussion

Regarding the magnitude of EU politicization in the six MAPLE countries (Fig. 3.1), we can observe substantial cross-national variation in terms of EU salience. On the one hand, as we anticipated, the EU has been a relatively salient dimension in the German parliament. The same can also be said for Greece and Ireland. In fact, it was in Greece that, on average, the EU was salient the most (20.2% of the speeches), followed by Germany (16.6%) and Ireland (14.3%). Moreover, it seems that the Eurozone crisis might have indeed contributed to a higher visibility of the EU in the national parliaments of Greece and Ireland. On the other hand, the salience of the EU has been comparatively low in the other three parliaments, particularly in the case of Portugal where, despite a slight increase after 2009, only 2.9% of the speeches, on average, mentioned an EU-related term. In this particular dimension of EU politicization, Belgium and Spain had on average 6.2 and 6.8% of EU speeches, respectively. Moreover, in the case of Spain, we can also observe a slight increase of EU salience after 2011.

Fig. 3.1

The Politicization of the EU per legislative term in six parliaments (EU salience indicated in the lines)

When it comes to contestation, different from the EU salience, the variation within countries seems more relevant than the differences between them. Nevertheless, on average, the EU has been a more contested topic in Greece (35.3), Belgium (31.8), and Portugal (28.3). Contrarily, the German parliament showed the lowest levels of EU contestation. On average, only 22.4% of the Bundestag speeches about the EU had a negative tone. It is very interesting to see that, as we anticipated, and in line with previous studies (e.g., Auel & Raunio, 2014b), with the exception of Greece, the EU contestation is higher in the parliaments with low EU salience and vice-versa. To put it in other words, apart from Greece, EU contestation has hardly been a prominent aspect in the plenary debates of the six countries. Still, we can observe an overtime increase in EU contestation in some of the countries, particularly in Portugal, Spain, and Greece.

The second aspect analysed in this chapter concerns the different EU dimensions debated in the parliaments. Once again, we can observe noticeable differences between the six countries (Fig. 3.2). In some cases, like Spain and Portugal, domesticated policies have been consistently the most salient dimension of the EU debated in the parliaments. On the contrary, in the other four countries, the debates about EU policies were consistently more frequent. The remaining dimensions of the EU have been much less debated in the plenary sessions. The only exceptions are the comparatively higher salience of constitutional structure debates in Portugal and, to a lesser extent, also in Germany.

Fig. 3.2

The salience of different EU dimensions in legislative debates

As we anticipated, the analysis confirmed the low salience of the membership dimension in the parliamentary debates of the six countries. This dimension was only relatively salient in some of the periods analysed in the Greek and German parliaments. In fact, in Greece, the results suggested that the high salience of the membership dimension (before 2003, 2012, and 2015) might be related to important EU events, namely the 5th EU enlargement and the Eurozone crisis. Overall, the low salience of this dimension, which is undoubtedly the most polarizing manifestation of the European integration process, might help us understand not only why the overall EU contestation is low in the institutional arena but also why the EU contestation is comparatively high in the Greek parliament.

When it comes to this dimension of membership, we can also observe some interesting differences between political parties (Figs. 3.3 and 3.4). Firstly, we can see that, particularly when looking at Greece, but also Belgium and Germany, the parties from the right were likelier, compared to the ones from the left, to discuss the membership dimension of the EU. Secondly, the discussion of the membership dimension, in Belgium and Germany, seems to be mainly ‘driven’ by the more Eurosceptic parties. Somehow differently, in Greece, the situation seems to have changed over time. While the Eurosceptic parties focused more on the membership dimension in the first legislative term analysed, the pro-EU parties started to emphasize more this dimension of the EU in the following plenary debates, making this change particularly evident after the beginning of the eurozone crisis. Finally, we can also observe a clear difference between pro- and anti-EU parties in Belgium when it comes to the type of policies discussed. More concretely, while the pro-EU parties in Belgium focus predominantly on debates about EU policies, their Eurosceptic counterparts prioritize the domesticated policies in their interventions.

Fig. 3.3

The salience of different EU dimensions in legislative debates by left and right parties

Fig. 3.4

The salience of different EU dimensions in legislative debates by Eurosceptic and Pro-EU parties

The remaining analyses made in this chapter focus on party differences. More concretely, we examine the politicization of the EU in national parliaments by making three distinctions, or categories, of parties (ideology, incumbency, and EU stance). The first category is whether a party is from the left or right (Fig. 3.5). When it comes to EU contestation, the only clear pattern that we can observe between left and right is in Ireland. More precisely, in the case of Ireland, the parties from the left have been consistently more negative towards the EU. Interestingly, at the same time, Irish parties from the right were the ones more likely to mention the EU in their speeches.

Fig. 3.5

The politicization of the EU in legislative debates by left and right parties (EU salience indicated in the lines)

In addition to that, it seems that, overall, the EU is indeed likelier to be more contested by left-wing parties, particularly after the Eurozone crisis. This pattern is particularly evident in the group of creditor countries. In Belgium and Germany, while the EU had been a more contested issue for the right in the first two legislative terms, the situation inverted in the remaining period of analysis. There are some possible explanations for the higher EU contestation levels from leftist parties. One is that, at least in this group of six EU countries, the extreme parties from the left are likelier to have better electoral success and, therefore, more visibility in national parliaments. Nevertheless, it is interesting to see that, overtime, our data shows that the EU has become increasingly more contested on the left side of the political spectrum.

The second distinction that we explore is whether a party is in government or not. As we saw, the literature suggests that government parties are likelier to discuss the EU in their plenary interventions. This was clearly not the case for Greece and Belgium, where the opposition has been, in a very consistent way, likelier to mention the EU in their speeches (Fig. 3.6). In other countries, government parties indeed seem likelier to talk about the EU, even though those differences seem relatively small and inconsistent. Once again, the most consistent patterns were found for EU contestation. The parties in the opposition have been more negative towards the EU in practically all the periods analysed. The only exceptions were a single legislative term in Belgium, Greece, and Spain.

Fig. 3.6

The politicization of the EU in legislative debates by government and opposition parties (EU salience indicated in the lines)

The third characteristic analysed is the parties’ position towards the EU (Fig. 3.7). In this case, we anticipated that the more Eurosceptic actors were also more likely to politicize the EU during parliamentary debates. With the exception of Spain, this expectation was confirmed in all countries. The difference is particularly clear in terms of EU contestation, with the Eurosceptic parties being much likelier to discuss the EU in a negative way. In the cases of Greece and Belgium, the Eurosceptic parties were also more likely to mention the EU in their speeches in comparison with the parties with a more pro-European stance.

Fig. 3.7

The politicization of the EU in legislative debates by Eurosceptic and Pro-EU parties

As we mentioned, the Spanish parliament is the only exception to this pattern of higher EU contestation from Eurosceptic parties. Interestingly, this was only the case in the three legislative terms after the Eurozone crisis. In this sense, it seems that this crisis had some impact on how—or by whom, to be more precise—the EU was politicized in the Spanish parliament. Nevertheless, in the last term, once again the Eurosceptic actors had a more negative tone when debating the EU.

Overall, when it comes to differences between parties, the clearest pattern identified in our dataset is that Eurosceptic parties were consistently likelier to contest the EU in their parliamentary speeches. This result, while not surprising, is an important validation of our measure of EU contestation. Still, the differences between parties were not as straightforward when it comes to EU salience. In fact, when looking at the institutional arena, salience, and contestation, differently from the media, seem to follow different logics. More concretely, the media logic of giving more salience to contested topics does not seem to apply to the plenary speeches (Table 3.3).

Finally, we investigate in this chapter the relationship between EU politicization and key party-level characteristics using linear regression analysis. The coefficients of the regressions, using EU salience and EU contestation as dependent variables, are plotted in Fig. 3.8. These plots refer to models⁸ 3 and 6, with the country controls omitted and 95% confidence intervals. As we can see, when it comes to the salience of the EU in the parliamentary speeches, the parties’ position on the EU issue is the only variable of interest with a statistically significant effect. Moreover, as we anticipated, the more Eurosceptic a party is, the likelier it is for the EU to be mentioned in that party’s parliamentary interventions. This variable was statistically significant (at the 0.001 level) in the three models.

Fig. 3.8

Coefficient plots from linear regression analyses on EU Salience and Contestation (models 3 and 6 from Table 3.4, in the Appendix), with omitted country dummies

Concerning the remaining explanatory variables, we did not find a statistically significant effect of ideology (left–right positioning) and incumbency (being in government) on the salience of the EU. The only other aspect in the models, besides EU position, with a significant effect on EU salience, concerns the period analysed. The models show that the salience of the EU has increased in the parliaments over time. Moreover, as we anticipated, the salience of the EU in parliamentary debates post-2009 was also higher, being the difference statistically significant at the 0.05 level.

When it comes to EU contestation, which refers to the proportion of EU speeches that were negative, we have found negative and statistically significant effects for all three main independent variables (EU position, Left/Right, and Government). The strongest predictor of EU contestation in parliamentary speeches was the party’s position on the EU, with the models showing that, unsurprisingly, the more Eurosceptic parties were also much likelier to contest the EU in their speeches. This is an anticipated finding that, nevertheless, further validates our automated measure of EU contestation.

The parties’ ideology, on the left–right scale, has also a statistically significant effect on the proportion of negative speeches towards the EU. More concretely, the levels of EU contestation increase significantly when the parties’ ideological positioning moves to the left. Contrarily, the parties on the right seem to debate the EU dimension with a more positive tone. When it comes to the effect of incumbency, as we anticipated, the EU dimension is more likely to be contested by the parties in the opposition.

Overall, in line with the existing literature, we found that parties’ position on the EU, and being in opposition, are both important predictors of EU contestation in the parliaments. Moreover, the contestation of the EU also seems to have been increasing over time, having been significantly higher in the period post Eurozone crisis. Considering that time is a good predictor of both EU salience and contestation, our analysis indeed suggests that the democratic legitimacy of the European integration has slightly increased over time, in the past two decades.

Conclusion

In this chapter, we offered an extensive analysis of the EU politicization in the institutional arena. Similar to previous assessments, our analysis confirmed the relatively low politicization of the EU in the institutional arena, a situation that even the eurozone crisis was not able to noticeably change. Despite important cross-national differences, the overall salience of the EU in the parliaments has been relatively stable and low. When it comes to the contestation dimension, while its variation seems more pronounced, the general tone towards the EU has only swung from very positive to less positive. In this sense, despite the intentions expressed in the Lisbon treaty, we cannot say that, perhaps with the exception of Greece, national parliaments have made a very noticeable contribution to reduce the democratic legitimacy of the European Union.

Considering the discussed determinants of EU politicization, and the reason why parties might strategically avoid politicizing the EU, it remains puzzling the reason why the Eurozone crisis did not lead to a more noticeable politicization of the EU in national parliaments, particularly when compared to the media. Factors such as the lower number/variety of active actors (compared to the media), the institutional obligation to a more ‘responsible’ (and less ‘responsive’) action, as well as the need to maintain a sense of legislative capacity (to avoid suggesting that their hands are tied), might have all contributed to this outcome. Furthermore, it is also possible that the existence of EU-specific committees has also contributed to this low salience of the EU in the plenary debates. Regardless of the reasons, this chapter highlights the importance of looking at each country individually to better understand the different nuances and implications of the EU politicization.

Nevertheless, our statistical analysis of EU politicization confirmed two important aspects regarding the party-level determinants of EU salience and contestation. On the one hand, the position of the parties in the EU issue was the strongest predictor of EU politicization, being statistically significant for both EU salience and contestation. On the other hand, the parties’ position on the left–right scale and being in the opposition had a significant effect only on contestation. More concretely, our analysis showed that, in the parliamentary speeches of the six countries concerned, the EU was likelier to be contested by the parties on the left and the parties in opposition.

Overall, this chapter confirmed the existence of strong cross-national variation at the institutional level. Our results highlight the importance of a less extensive and more in-depth analysis of the EU politicization, as well as suggesting the importance of the arena where politicization is the measure. Using a different methodological approach, the next chapter offers a more detailed picture of how the EU was politicized in the six countries and the similarities between the institutional and the intermediary arena of political communication.

Appendix

See (Table 3.2, Table 3.3, and Table 3.4)

Table 3.2 List of the expressions used to identify EU articles

Table 3.3 Summary of descriptive statistics of the variables analyzed

Table 3.4 OLS regression analysis to identify the determinants EU salience and EU contestation in the parliamentary speeches of the six MAPLE countries