Keywords

Introduction

In this chapter, we bring together the media, parliamentary and voting data to provide a comprehensive picture of the role of EU politicisation for Portuguese voting. The Portuguese case-study will contribute to the overall goal of the book, namely to understand whether EU issue voting is occurring and national channels of representation are serving as accountability mechanisms for the process of EU integration. The volume begins by setting the stage, namely by providing the trends on EU politicisation in the media and parliamentary debates in Europe, both before and after the Eurozone crisis. Then, Chapters 5 and 6 establish the existence of EU issue voting in all countries, demonstrate that it is magnified by increases in media salience and parliamentary debates’ negative tone, and show the prominence of left–right positioning over other issues in the four bailout countries considered, namely Greece, Ireland, Portugal and Spain.

Yet, there are still unanswered questions which only an in-depth analysis of the Portuguese case-study can provide, and that constitute the focus of this chapter. First, we present the salience and tone of the EU issue in media and parliamentary debates in Portugal in a longitudinal perspective. This will allow us to understand how present the EU was in the media, compared to past elections, and thus give a proper sense of how important it was in the 2019 election context. Then, comparing the EU issue’s relative importance, with other political issues, which is the goal of this chapter, we are in effect benchmarking our findings on EU issue voting.

Portugal is an interesting case, to analyse the phenomenon of EU politicisation and its consequences, for several reasons. First, because it has been seen as a country of euroenthusiasts, both in terms of its political elites, as well as its citizens (Llamazares & Gramacho, 2007; Jerez-Mir et al., 2009; Verney, 2011). While this relative consensus may caution against politicisation, research has demonstrated that, in fact, not only the EU was politicised, especially on the Left of the party spectrum, but it was also an explanatory factor of voting behaviour for extreme-Left positioned citizens (Lobo, 2003, 2021). Second, because Portugal was one of the countries at the epicentre of the Eurozone crisis. Portugal had to ask for a bailout in 2011, which lasted until 2014. The bailout brought with it stringent fiscal policies, with governments agreeing to harsh cuts both in public sector wages, in pensions and other welfare subsidies, as well as tax increases. The austerity which ensued from the bailout raised the profile of the EU in the country, and may have made it a more relevant issue for voting. Third, because the data we collected pertain to the 2019 legislative election, held on the 6th October, which can be considered a post-bailout election, and where the exacerbated role the EU may have played during the bailout may have subsided. Thus, any effects which are found now may be considered more long-lasting, independent of the crisis.

The Chapter is organised in the following way: first we start by presenting a brief overview of the literature on parties, voting and the EU in Portugal. Then, we describe briefly the context of media and parliamentary debates to understand the data we present. Next, we explain the general political context of the 2019 elections and the data collected in relation to Público and Diário de Noticias, two mainstream newspapers, both in terms of salience and tone, from 2002 to 2019 as well as the data on parliamentary parties’ salience and tone, during the same time-frame. It is relevant to discuss both these arenas as different forums where information about the EU may have been communicated from parties to citizens. Also, as shown in Chapter 4, these arenas differ in communicating the EU.

Then, we present the analysis of EU issue voting in the 2019 legislative elections. We consider the degree to which the EU issue explains the vote for each major party on the Left and on the Right, using multinominal regression and presenting Average Marginal Effects of EU issue voting compared to other issues, in two different models of voting behaviour. The first includes socio-demographic controls, ideology and different issues, including the EU issue. The second includes all the variables in Model 1, as well as the short-term variables of attitudes towards the leader of the party voted for and economic perceptions. In effect, we will be able to determine the relative importance of the EU in the context of other issues, which have been deemed important for political debate in Portugal.

Parties, Voting and the EU in Portugal

The Portuguese party system has been characterised by “limited EU contestation” (Taggart & Szczerbiak, 2008) with Euroscepticism being politicised mainly on the Left (Freire & Teperoglu, 2007; Sanches & Santana-Pereira, 2010; Santana-Pereira & Fernandes, 2014). Indeed, the European cleavage was, since democratisation, contained in a larger “regime cleavage” and served to distinguish the Communist Party from the other parties with parliamentary seats. Namely, the Socialists, PS, the centre-right PSD, as well as the conservative CDS considered, that European integration would be useful for the consolidation of a liberal democracy in Portugal (Pinto, 2011). Contrarily, the Communists considered that EU membership would be a way of ensuring the diffusion of right-wing neoliberal policies in Portugal. Then, in 1999, another left-wing party gained access to Parliament: the Bloco de Esquerda (Left Block), which joined the Communists in adopting an Eurosceptic stance (Lobo & Magalhães, 2011). The BE concurred on the negative economic consequences for Portugal of the EU, but had a pro-EU stance in other dimensions of the European integration, namely on the benefits of the EU project itself. These are the only two parties which have been systematically Eurosceptic in the Portuguese party system. On the Right, the CDS-PP flirted with Euroscepticism in the late 1990s as a way to distinguish itself from the centre-right governing party, the PSD. Yet, since the early 2000s, it became a steady coalition partner for that party and abandoned its Eurosceptic agenda (Lobo, 2003; Sanches & Santana-Pereira, 2010).

Since the onset of the Eurozone crisis there have been notable changes in the party system. Firstly, there has been a tendency towards a decline of the strength of the two main parties, the PS and the PSD. While between 1987 and 2005 the sum of their votes was on average 76%, from 2009 to 2022 it has decreased to 68%. This value is slightly inflated considering that in 2015, we also added the votes of the CDS, taking into account that PSD and CDS formed a pre-electoral coalition in that year.

Secondly, there was a change in the dynamics of the party system following the crisis (Lobo, 2021), with greater bipolarisation. In 2015, the Socialist party, which came second in the elections, decided to form a majority alliance with the parties to its left, the Left Block (BE) and the Communists (PCP). This was the first time such an alliance was held in the history of Portuguese democracy, and it took the form of a coalition of parliamentary incidence, with the smaller parties not taking any seats in government. In part, it had not happened until then due to ideological differences between the parties, including their position on Europe, detailed above. The coalition lasted its full mandate, until 2019, but the parties presented themselves independently to the election. This is an important election to study from the perspective of media, parliamentary debates and votes, since the unprecedented alliance of these parties, which lasted the full mandate 2015–2019, may have mitigated the importance of the EU issue across the different forums and even for voting behaviour.

Thirdly, since 2019, there has been fragmentation on the Right. In that election, there were three new entrants into Parliament, two of them on the Right: one MP each from an extreme-right party Chega, and a liberal party, Iniciativa Liberal. The entry of the far-right Chega into Parliament constituted an important watershed, as Portugal was one of the few remaining countries of Europe without an extreme-right populist party (Mendes & Dennison, 2021). The third party to enter Parliament was Livre, a left-libertarian party, which also elected one MP. In the 2022 elections, both Chega and IL increased their vote substantially, with Chega becoming the third most-voted party in Parliament, winning 7% of the votes and 12 MPs. These parties on the Right have different positions regarding the EU. While Chega adopted in 2019 an “Europe of Nations” Eurosceptic position, IL was clearly pro-EU. Taken together, it seems that the Eurozone crisis did have some important consequences for the party system in terms of dynamics, as well as its format.

In contrast with research on party positions, there has been relatively less research on the importance of the EU for individual attitudes and political behaviour, in particular. In terms of attitudes, the Portuguese electorate initially combined a very positive outlook on the EU with a relative lack of knowledge and interest. Moreover, largely positive attitudes did not translate into electoral participation for EP elections (Lobo, 2011). Indeed, Portugal has one of the lowest levels of participation in EP elections, even when we consider EP elections from 2004 onwards, which already include the Central European countries. Moreover, in 2019, the first time the EP elections reached an overall 50% turnout, in Portugal only 37% of voters participated in the elections. When it comes to European attitudes, it was systematically found that support for the EU in Portugal was rather instrumental, dependent more on economic benefits than on political values of membership (Lobo, 2011).

Concerning what explains support for the EU in Portugal, satisfaction with democracy was the most important variable in explaining support for the EU, followed by voting for the Communist Party (Lobo, 2003). More recently, Freire et al. (2014) show that the onset of the Great Recession led to a strong growth in Euroscepticism at the voter level in Portugal, as occurred in other bailout countries. The authors found that having an extreme-left or an extreme-right positioning was predictive of Eurosceptic attitudes, in line with the findings by Santana and Rama (2018). In addition, even after controlling for all the major factors of Euroscepticism, attitudes towards the Troika agreement and debt renegotiation had a significant impact on voter’s support for the EU.

Teperoglu and Belchior (2020) find that in Portugal, at the peak of the crisis, self-placement on the centre-left and, to a lesser extent, on the extreme left was a significant determinant of Eurosceptic stances, but this effect had lost significance by 2018. On his part, Lisi (2020) shows that extreme-Left voters are the most Eurosceptic; negative economic perceptions fuel Euroscepticism and those who tend to trust national institutions also trust the EU to a greater extent. In what concerns voting, there is a difference between the two main parties (PS and PSD), whose sympathisers are clearly pro-European, and challenger parties. As expected, both BE and PCP sympathisers are more Eurosceptic, in a significant way. Moreover, for explaining the vote, the EU issue was significant in voting in the 2014 EP elections (Freire & Santana-Pereira, 2015). Overall, the studies have shown that Euroscepticism is located on the left of the electorate, and that the crisis sharpened the existing differences. We turn now to the analysis of the EU across the two different contexts: media and parliamentary debates.

Studying the EU Across Different Forums: The Case of Portugal

As has been explained above, in this Chapter, we examine different kinds of data, namely media articles and parliamentary debates to contextualise voting. In this section we seek to present a brief literature review on each of these different political arenas research in Portugal to better understand the data included in the chapter. Concerning the media landscape, Hallin and Mancini categorised Europe’s media systems (2004, 2012), following a number of structural criteria. According to them, Portugal belongs to the polarised-pluralist type of media system, alongside Spain, Greece and Italy. This model is defined by a weak, underfinanced media market, political control of the media, and state intervention in the media as owner, regulator and financial backer (Santana-Pereira, 2016).

Indeed, Portugal may be considered a case where campaigning is permanent, and it occurs through the various mass media, especially television and newspapers (Santana-Pereira, 2016). The way that “permanent” campaigning happens in Portugal is through the role that party politicians taken on as pundits both in television and the main newspapers (Figueiras, 2011, 2019). Figueiras has shown that the time/space dedicated to punditry has increased since the 1980s in Portuguese mass media (2011). Yet, according to Silva et al. (2017), the instrumentalization of the media has been mitigated in Portugal, relative to other Southern European countries. This is due to the fact that there is less differentiation between mainstream parties, as well as the political professionalism of journalists in general. Indeed, repeated surveys show that the media, both television and newspapers are highly trusted in Portugal (Newman et al., 2019).

Within this context of high interpenetration between politicians and mass media, there is little information about the kinds of topics which are discussed during campaigns in a systematic fashion. Concerning the EU topic, it was found, considering the 2009 EP elections, that the salience of the EU during EP elections was the highest in Portugal, of the 13 countries considered. In terms of tone, whereas the majority of countries’ media had a positive tone towards the EU, in Portugal it was slightly negative (Stromback et al., 2011). In the same volume, Jalali and Silva (2011), find that there are differences between the way in which parties and media politicise the EU, as well as differences between parties. They find that the media is significantly less preoccupied with Europe—in terms of issues and themes, if not actors—than the parties. Further, they find that it is government parties that focus on the EU issue, whereas opposition parties, of the left and the right, focus mainly on national issues. We now turn to the analysis of the 2019 election, and analyse the media and parliamentary data in a longitudinal fashion.

The 2019 Election Campaign: Media and Parliamentary Data

The 2019 election followed the end of the full mandate of a minority Socialist government, which had the support from the Left Block and the Communist Party. Following the 2015 elections, which were the first post-bailout elections, the right-wing party, PSD, won the election but without a majority. However, rather than supporting the PSD’s executive, the Socialists decided to form a coalition of parliamentary incidence with the two small parties on its left. This was the first time that these parties were able to form a coalition, which was labelled “geringonça”. The mandate between 2015 and 2019 proved to be politically stable. During that period, the indicators of support for democracy as well as trust in government improved quite substantially. With improving economic indicators, the major beneficiary of this government mandate was the Socialist party, which saw its vote increase from 32.31% to 36.34%, while both the PCP and the BE saw their vote decrease slightly (Fernandes & Magalhães, 2019). Yet, despite this result, following the 2019 elections, the Socialists decided not to re-enact their coalition of parliamentary incidence. Instead, they formed a minority government without any parliamentary agreements (Jalali et al., 2020). This election also saw the entry into Parliament of three new parties: on the far-right, Chega, on the liberal right, Iniciativa Liberal, and, on the Left, Livre, each elected one MP to the Assembleia da República, winning seats in the Lisbon electoral district, the largest in the country. Thus, while confirming the resilience of votes for the centrist parties, the PS and the PSD, 2019 also saw the entry of three new parties to Parliament, which signalled future party system fragmentation.

Next, we present data on the politicisation of the EU in media and parliamentary debates for the period 2002–2019. It is relevant to discuss both these arenas as different forums where information about the EU may have been communicated from parties to citizens. “Politicisation” has been defined as a process whereby a collective decision generates disputes, and wherein the audiences of those disputes gradually expand (Schmitter, 1969). It refers to “an increase in polarization of opinions, interests or values and the extent to which they are publicly advanced towards the process of policy formulation” (De Wilde, 2011, 559). In our research, politicisation has been operationalised in two dimensions: salience and polarisation (Silva et al., 2022). Salience is measured through the number of articles/speeches which mention the EU in a significant way (in the title or in the article body for media articles, and in speeches for parliamentary debates) as a proportion of the total number of articles/speeches. Polarisation is harder to measure. For our purposes, we opted for presenting the tone of the articles/speeches which mention the EU. In the case of media, this measure refers to an average of two measures obtained in an automated fashion. Namely, it combines a measure of the sentiment of the title of the article, and a measure of the average sentiments of the EU sentences in the article itself. In the case of parliamentary debates, this measure refers to the average sentiment in all the EU sentences uttered in a given speech. Thus, once the EU sentences have been identified, a sentiment score is calculated for each of them, after having been translated into English, and the tone measure is the average sentiment score within those speeches (see also Chapter 3).

In Fig. 11.1, we present data from the two main daily newspapers, Público and Diário de Notícias, which are traditionally associated respectively with the centre-left and the centre-right of the Portuguese party spectrum. For each time point, the average of salience and tone is presented for the relevant newspaper articles in the thirty days before the election. Rather than present only the 2019 data, we opted to present data longitudinally from 2002 to 2019, which helps us to understand the specificities of our election of interest. Thus, the data below present EU salience and tone in Portuguese mainstream media from 2002 to 2019.

Fig. 11.1
A combination chart illustrates the variations in the E U salience and the E U tone based on two main daily Portuguese newspapers Público and Diário de Notícias from the years 2002 to 2019.

EU politicisation in the Portuguese media, 2002–2019 (Source MAPLE data)

Full size image

Firstly, as has been noted elsewhere (Silva et al., 2022), there was a sharp increase in media salience to the EU following the onset of the Eurozone crisis, which has not abated. Thus, for both newspapers, the period post-2011 shows greater EU salience than pre-2011. Despite this trend, there are some differences between newspapers. Namely, Público consistently mentioned the EU to a greater extent than Diário de Noticias, and this was especially the case in 2011, the election which followed the bailout agreement in Portugal.

When we consider tone, the 2002–2015 trend is the following: there has been a decline in the positive tone of articles, signalling an increase in negativity of articles, from 2005 onwards in both newspapers. When we consider tone, we also detect differences between the newspapers. Whereas centre-left Público has a steady decline in tone from 2002 onwards, in centre-right Diário de Noticias, the tone becomes more positive from 2002 to 2009, and then drops precipitously until 2015.

When we focus on tone in 2019, we see that contrary to past trends in both newspapers, tone improves in 2019, to pre-crisis levels (2009) both in Público and Diário de Notícias. Thus, it seems that although attention to the EU suffered a dramatic shift post-2009 for both newspapers, which has not been undone in 2019, in terms of tone, there was an improvement in the latest election. There could be several reasons for this improvement in tone. First, the simple fact that the bailout has ended, and that Portugal was able to meet its public finances commitments in the EU after 2014, eased relations with the EU. Second, the government’s Finance Minister, Mário Centeno, became President of Ecofin, which may have contributed to more positive-toned articles mentioning the EU. Thirdly, as explained above, the main Eurosceptic parties in Portugal, PCP and BE were supporting the government, and thus less likely to effectively politicise the EU. The data on parliamentary debates, that we discuss next, indeed confirm this.

Unlike the media data, where each time point represents data collected one month before the election, we were able to collect all speeches for the entire year for parliamentary debates. Thus, the parliamentary debates dataset includes all plenary speeches which were uttered from 2002 to 2019.

Considering salience, we note that EU salience in plenary speeches is low, and does not increase dramatically since 2009 (Fig. 11.2). Yet, the differences between parties increase following the crisis. Moreover, from 2011 onwards, the mainstream parties tend to distinguish themselves. The salience given by the Socialist party has a higher increase, vis-à-vis the salience attributed by the PSD, independent of the former being in opposition (2011–2015) or in government (2015–2019). The Communist party exhibits a stable pattern, with higher salience attributed at election times, while the Left Block gave EU issues more salience during the bailout period (2011–2014), than since they supported the PS minority government (2015–2019). On the Right, the CDS-PP seems to follow closely the PSD in the salience attributed to the EU. They were coalition partners from 2011–2015, and incumbency increased the salience attributed to the EU.

Fig. 11.2
A multiline chart demonstrates the variations in the E U salience in parliamentary debates per Portuguese parties B E, P C P, P S, P S D, C D S P P, and P A N from the years 1999 to 2019.

EU salience in parliamentary debates per Portuguese Party (Source MAPLE data)

Full size image

Figure 11.3 presents the tone employed by parliamentary parties from 2002 to 2019. Contrary to salience, where parties hardly distinguished themselves from each other until the onset of the crisis, in the case of tone we see differences for the whole period. Concerning the mainstream incumbent parties, the Socialist party tends to be less positive when in opposition (2002–2004; 2011–2015) than when in government (1999–2001; 2005–2009; 2015–2019). In particular, the most positive tone coincides with the year in which the PS won its first absolute majority (2005), and on average remaining until 2011 with a more positive tone than the mainstream opposition party, the PSD. When the latter party formed government in 2011 to administer the bailout, and governed until 2015, its tone was more positive than the Socialists, which were at its lowest for the period analysed. Once the PS returned to power in 2015, its tone towards the EU becomes more positive. On the right, the PSD tone was most positive when the party was in government (2002–2004; 2011–2014), and less so when in opposition, in 2014–15, only to recover during the more recent period. Considering the smaller parties, the CDS-PP seems to emulate the PSD trends. On the Left, as we would expect, there are more differences. Since joining the euro, in 2001, the Communist party as well as the Left Block have had a rather more negative tone regarding the EU than the Socialists. For the Communists, the tone decreases from 2006 to 2015, where it reaches its lowest point. The Left Block follows this trend too, with low and declining levels of tone from 2006 to 2012. Forming the left alliance with the Socialists seems to have had some impact for these two parties. Both parties, especially the Communists, saw increases in tone, from 2015 onwards. In that period, the Communists’ tone was almost identical to the PS, something that had not happened since 2001.

Fig. 11.3
A multiline chart demonstrates the variations in the E U tone in parliamentary debates per Portuguese parties B E, P C P, P S, P S D, C D S P P, and P A N from the years 1999 to 2019.

EU tone in parliamentary debates per Portuguese Party (Source MAPLE data)

Full size image

Therefore, taken together, we observe that, in terms of EU politicisation, 2019 saw the relative salience in the media, but slightly less polarisation in the newspapers during the campaign. A similar trend was also observed in the legislative term in Parliament (2015–2019), where the tone of opposition parties was on average much more positive than during the previous years. All in all, we can consider that there was a depoliticisation of the EU in these two political forums in 2019, compared to the previous period in the Assembleia da República. Thus, this context is one where, despite more awareness of the EU, due to its salience, it is a time of less contestation, especially due to the fact that the Communists and the Left Block were supporting the government. This leads us to expect that EU issue voting may not be very significant in these elections. Interestingly, this, however, doesn’t seem to be confirmed in our analysis as we will see in the next section.

Benchmarking EU Issue Voting in Portugal

The survey employed is a representative two-wave panel online survey with a sample of 1540 in the first wave and 1608 respondents in the second one. We are using the second wave, post-election data, which was collected between 7th October and 30th November 2019. The panel provider was able to fulfil a crossed quota of gender (2 categories), age (3 categories) and education (3 categories), using the 2011 census as the matrix to build the sample. A model of voting behaviour was built which includes socio-demographic controls, ideological self-placement, political issues, including the EU, leader barometers and perceptions of the economy to explain the vote for each main party that won seats in 2019. We only included in the analysis parties where at least 40 respondents stated they had voted for it. Thus, our analysis includes the Left Block (BE), Communists (PCP), Animal Party (PAN) and Socialists on the Left, and PSD, CDS-PP, and Chega on the Right.

Our dependent variable is vote recall, which is a categorical variable. We employ multinomial regression, with the Socialists as the baseline reference. As the coefficients of multinomial models are complex to interpret and depend on the chosen baseline outcome, the results are here presented, in Fig. 11.4, by plotting average marginal effects of the main variables of interest. The regression results are presented in the Appendix: Table 11.1. Average Marginal Effects of the key issue variables are presented, to contrast the importance of EU issue voting among all parties, and also to contrast the relative importance of EU and other political issues.

Fig. 11.4
Seven forest plot graphs demonstrate the average marginal effect of voting for each main Portuguese party B E, P C P, P S, P S D, C D S P P, and P A N.

AMEs of voting for each of the main Portuguese parties (Note Average marginal effects and 95% confidence intervals, based on the results reported in Model 1 in Appendix: Table 11.1)

Full size image

Our main independent variable to measure position on the EU issue is the following variable: “Some people believe that the process of European integration should move forward to the creation of the United States of Europe. Others believe that the European Union should be dissolved in order to return to a situation in which states are fully sovereign. In which point of the following scale would you place yourself?” (10-point scale from completely agree with dissolution to completely agree with a United States of Europe).

Other issue variables, salient due to their relevance for the 2019 campaign, were included to benchmark the importance of the EU issue. Firstly, attitudes towards public services and social protection, measured with the following question: “Using a scale from 0 to 10, where 0 means “We should improve public service and social protection even if it means a tax increase” and 10 means “we should reduce taxes even if it means reducing public service and benefits, where would you stand?” Secondly, attitudes towards equality in redistribution, using the following item: “And now using the same scale from 0 to 10, where 0 means “there should be more equality in income distribution” and 10 means “there should be more incentives for individual initiative” where would you stand? Thirdly, we measure attitudes towards the national health system in the following way: “Finally, using the same scale from 0 to 10, where 0 means “the national health system should be privately controlled” and 10 means “the national health system should be state controlled”, where would you stand?

Simple leader barometers, which we have included in the survey, and have been shown to be highly correlated with party choice, were also included in the regression analysis. Also, sociotropic retrospective economic perceptions were included in a different model. Therefore, we will show the results of two models, first without leaders and economy, and then including these two variables, to understand if the significance of EU issue voting resists the inclusion of these important short-term variables. Another reason to have two models is due to the fact that we did not include a question about attitudes towards André Ventura, the Chega leader. For this party, its results are only presented in the first model.

Socio-political control variables were also included to understand the relative importance of each in explaining the vote for each party. Namely, age (3 categories), gender (2 categories), education (3 categories), religiosity (4-point scale) and ideology (11-point scale). All independent variables were standardised to vary between 0 and 1, for the sake of results interpretation.

The regression results presented in the Appendix: Table 11.1 show that concerning socio-political controls, on the left, BE voters tend to be younger than the Socialists, whereas the Communists are more unionised. Both BE and Communists are significantly less religious than the Socialists. On the Right, men and individuals with higher education are likelier to vote for CDS and PSD, compared to PS.

The general left–right continuum performs well in the model, with those who voted BE and Communists being significantly more to the left of the Socialists, and PSD and CDS voters to the Right. PAN voters do not distinguish themselves ideologically from PS. Chega voters in this sample are not significantly different from the socialists in terms of ideological self-placement, but caution should be had relative to this party as the number of respondents in the sample which stated they voted for that party is very low.

The results in Fig. 11.4 present the Average Marginal Effects for the EU issue as well as the other socio-economic issues and left–right. When examining the recalled vote after the election, there is a significant negative effect for the Communist party with voters 10 percentage points more likely to vote for this party when they oppose the EU than when they believe the EU should move forward to the United States of Europe. This effect is significant, as can be seen in Appendix: Table 11.1 in both models, with and without short-term variables. BE voters are also more likely to choose this party if they oppose the EU, while PS and PSD voters are, on the contrary, more likely to choose those parties if they support the furthering of EU integration. Yet, the relationships between the latter three parties are not significant, with the exception of the PSD.

Concerning the other political issues, being in favour of state control of the national health system has a significant positive effect in voting for Communists, while the opposite occurs for the CDS-PP and the PSD. The impact of the other two socio-political issues, namely improving public services and defending income redistribution is not significant. The political issue effects on voting for the Communists, the PSD and BE, with PS as the baseline, remain significant even after the inclusion of short-term variables, leader barometers and economic perceptions.

Conclusions

In this Chapter, we analysed trends in media and parliamentary politicisation of the EU in Portugal, as well as its consequences for voting behaviour. In order to do so, we employed unique data collected longitudinally, from 2002 to 2019, as well as panel survey data collected after the October 2019 legislative elections.

We showed that, since the onset of the crisis, two periods can be distinguished in terms of EU issue salience and tone both in the media, and in parliamentary debates. From 2009 to 2015, the whole crisis/bailout period, salience has increased and its tone has tended to become more negative in the mainstream news media. Then, from 2015, salience does not decline to pre-2009 levels, both in the media and parliamentary debates. Yet, the tone in the mainstream media, as well as in parliamentary debates improved in that period.

Thus, the parliamentary parties, in the mandate 2015–2019, did not really differentiate themselves in Europe. This may be a reflection of the fact that the main Eurosceptic parties, the Left Block and the Communists, had formally agreed to support the minority PS government, for the first time in 40 years, from 2015 to 2019.

We then turned to the analysis of the survey data. We saw that the EU issue explains the vote for the Communists and the PSD in 2019, relative to other factors. With the exception of Health, we also saw that most of the other political issue variables included, namely those relating to the welfare state, do not increase the likelihood to vote either for left or right parties. We also ran the models that included leader barometers and the economy and the results remained the same.

Overall, there is a continuity of the eurosceptic nature of the PCP electorate despite the fact that during the  2015–2019 left-wing alliance government, the party distinguished itself less from the Socialists in parliamentary tone regarding the EU. The results also show that the PSD voters are more euroenthusiastic about advancing EU integration than the PS. This is important as the PSD is the main opposition party, and it shows that the EU issue is not only important for voters on the extremes of the party system (Communists), but also for mainstream voters, such as those who vote PSD. Overall, the chapter shows that EU issue voting matters and its impact is larger than that of other comparably salient issues, which signals the importance that national institutions can have for the legitimization of the European Union.