Walgrave, S., van Erkel, P., Jennart, I., Celis, K., Deschouwer, K., Marien, S., Pilet, J. B., Rihoux, B., Van Haute, E., Van Ingelgom, V., Baudewyns, P., Kern, A. & Lefevere, J. (2020) RepResent longitudinal and cross-sectional electoral survey 2019. DANS. doi: 10.17026/dans-xe8-7t78.
-
1 For an overview of the distributions, see Walgrave et al. (2020). Weights were provided correcting for socio-demos (age, gender and education) and vote choice.
On 26 May 2019, citizens across Belgium went (in the context of compulsory voting) to the polling stations to vote in the third, so-called mother of all elections, casting votes for the federal, regional (i.e. Flanders, Wallonia or Brussels) and European parliament on the same day. Remarkably, many Belgian voters seized this opportunity to send a strong signal to the major governing parties, voting for one of the radical parties on the two ends of the ideological spectrum – the radical right Vlaams Belang (VB) (second party with 18.7% of the votes in Flanders) and the radical left PTB-PVDA (5.6% in Flanders, 13.7% in Wallonia). In addition, in Flanders, the first party was the N-VA with 24.8% and in Wallonia it was the PS with 26.1%. Such divided results had lasting effects, as they led to a new period of impossibility to form a majoritarian governmental coalition between the Flemish nationalists (N-VA) and the French-speaking social democrats (PS): the first being supposed to give more leeway on social reforms and the second on the transfer of federal competences to the regional level.
Even more generally, the outcome of the 2019 elections saw a substantial shift in party preferences. For instance, in the federal parliament 34 out of the 150 seats changed party ownership, marking the 2019 elections as the most volatile in Belgium post-World War II, at least if we look at the so-called nett volatility. It has to be kept in mind, however, that not only the electoral outcome was remarkable, but also that the months before 26 May were out of the ordinary, as the 2019 elections followed a turbulent political period. Earlier, in December 2018, the largest party in the federal government, the N-VA, resigned from government after a dispute about the signing of the Global Compact of Migration – also known as the Marrakesh pact – leaving the other government parties to form a new minority caretaker government.
The aim of this special issue is to take stock of what happened on 26 May 2019 and to explain the vote choice in Flanders and Wallonia. Who were the voters that voted for the populist radical left and right parties and what was their motivation? Were these voters – as was sometimes argued afterwards – driven by a resentment towards the traditional political parties and a request to reform the political system? Or were there ideological motivations underlying the vote shifts as well? And more generally, what was the role of ideology and the salience of issues such as immigration, economic situation or democratic reforms on the vote shifts in the two regions?
To answer these questions, most articles in this special issue use the same dataset: the RepResent Voter Panel Survey 2019 (Walgrave et al., 2020), which was conducted by the EOS RepResent Consortium – a collaboration between five universities (University of Antwerp, Vrije Universiteit Brussel, KU Leuven, Université libre de Bruxelles and UC Louvain) – in order to capture respondents’ electoral and political behaviour and attitudes. The study consists of pre- and post-electoral waves around the elections of 26 May 2019. The CAWI (Computer Assisted Web Interviewing) questionnaires were distributed online by Kantar TNS to Kantar’s own panel, as well as panels from other online companies such as Dynata. The target was a nett sample that would match the distribution on gender, age and education for the voting-age population in their respective regions. Nevertheless, due to non-response, the final samples differ from the target population distributions slightly; the higher educated are somewhat overrepresented just as the age group between 45 and 65.1xFor an overview of the distributions, see Walgrave et al. (2020). Weights were provided correcting for socio-demos (age, gender and education) and vote choice.
The first, pre-electoral wave ran from 5 April to 21 May 2019 – although 99% of the respondents were interviewed before May 6 – while the second post-electoral wave took place between 28 May and 18 June 2019. A total of 7,351 respondents took part in the first wave, of which 3,298 resided in Flanders, 3,025 in Wallonia and 1,028 in Brussels. Of those respondents, 3,917 returned for the second wave, which corresponds to a response rate of 53.3%; 1,978 respondents in the second wave resided in Flanders (60.0% response rate), 1,429 in Wallonia (47.2% response rate) and 510 in Brussels (49.6% response rate).
Based on this dataset, the articles in this special issue all revolve around the question of how to explain the vote choice of (specific groups of) voters. Generally speaking, these explanations fall into two broad categories that can be described as ideological on the one hand and emotional on the other. The latter category of reasons is expected to be largely determined by (dis-)satisfaction and resentment, being either operationalised by protest attitudes or, as in the third contribution, calls for institutional reforms.
In the first contribution, Ine Goovaerts, Anna Kern, Emilie van Haute and Sofie Marien study how protest attitudes and ideological considerations affected the vote for the three populist radical parties: VB on the extreme right end and PVDA on the extreme left end of the ideological spectrum in Flanders, as well as PTB on the extreme left end in Wallonia. They demonstrate that both types of determinants played a role when distinguishing between the voters of these parties and the electorates of mainstream political parties. However, when comparing voters of populist radical parties with voters who opted for mainstream parties that represent the closest alternative (i.e. N-VA for VB voters, sp.a for PVDA voters and PS for PTB voters), the study shows that it is primarily protest attitudes that distinguish these electorates from each other.
The second contribution to this special issue by Marta Gallina, Pierre Baudewyns and Jonas Lefevere also dedicates its attention to the support for the three populist radical parties. However, Gallina et al. pay special attention to the role of political sophistication. More specifically, they investigate to what degree political sophistication moderates the effects of both issue considerations and populist-related motivations such as political distrust, dissatisfaction with representative democracy and populist attitudes. They find, first of all, that the common image of voters of populist parties being rather unsophisticated citizens cannot be confirmed, as lower levels of political sophistication do not increase the probability of a vote for populist radical parties. Furthermore, they demonstrate that the effects of both issue considerations and populist-related motivations are indeed moderated by political sophistication, although this moderation is not uniform across parties. Higher political sophistication enhances the importance of populist-related motivations in the vote choice, although mostly for PTB and VB voters and less for PVDA voters. Also, for issue considerations, a moderation effect was found, with issue considerations playing a stronger role for more sophisticated voters; however, this only applies to VB voters.
The findings of the first two articles complement each other in the sense that we can conclude that the vote for populist radical parties is driven by both protest/populist attitudes and issue/ideological considerations, but that the extent to which these two explanations play a role is in part moderated by the political sophistication of the voters and is also not uniform across the different populist radical parties; such a result supports a more differentiated/nuanced use of the encompassing term/label of ‘populism’.
In the third contribution to this special issue, Lisa Van Dijk, Thomas Legein, Jean-Benoit Pilet and Sofie Marien add another explanatory factor to the equation that has so far been neglected, as they focus on demands for institutional reforms of representative democracy as a potential drive for the populist radical vote choice. Voters of populist parties’ support for different kinds of reforms (direct democracy, deliberative democracy, technocracy and limiting elected politicians’ prerogatives) are studied in conjunction and the authors confirm that depending on left and right positioning, voters actually differ in the kind of reforms that they perceive as desirable. While, in comparison to voters of non-populist parties, both VB and PVDA-PTB voters are more likely to support the idea of limiting the prerogatives of politicians and introducing binding referendums and participatory budgeting, their attitudes towards consultative reforms (both consultative referendums and citizens’ forums) as well as technocratic reforms differ. Whereas VB voters are actually indistinguishable from mainstream party voters regarding these reforms, PVDA-PTB voters are more inclined to support them.
Instead of focusing on voters of populist radical parties, Stefaan Walgrave, Patrick van Erkel, Isaïa Jennart, Jonas Lefevere and Pierre Baudewyns scrutinise another specific part of the electorate in the fourth contribution to this special issue, namely voters who combine a right-wing position on economic issues with a left-wing position on cultural issues and the other way around (i.e. left-wing position on economic issues and right-wing position on cultural issues). These two groups of voters are special, because in both the Flemish and the Walloon party systems no political party combines these positions. These atypical voters encompass about one fourth of the respondents in their Flemish sample and about one third in their Francophone sample. Since these ‘political orphans’ have no ‘fully matching’ party in the electoral competition, they need to prioritise either the cultural or the economic dimension in order to choose a political party. This study illustrates how the salience of those two dimensions explains whether these ‘unserved voters’ ultimately end up voting for a right-wing or a left-wing party. The authors furthermore demonstrate that this mechanism also holds for those voters who are not ‘underserved’ but rather ‘overserved’ in the sense that there are several parties which represent their policy preferences. As the study shows, these voters also rely on the salience of the two dimensions in order to make their vote choice and select the party that is closest to them on their more salient dimension.
Finally, Caroline Close and Emilie van Haute address another factor that drives vote choice: citizens’ emotions towards politics. Thereby, they focus on the link between several emotions (e.g. anger, fear, hope, joy) and all major parties in both Flanders and Wallonia. The authors, first of all, find that the emotional register of voters towards politics is at least two-dimensional, with a separate positive and negative dimension. These emotional dimensions have a significant and substantial effect on the vote choice. The study shows for instance that negative feelings, particularly anger and fear, represent major drivers for the support of VB, but that negative emotions in combination with a lack of positive ones can also prompt support for radical left parties, although to a lesser extent. Overall, Close and van Haute demonstrate that emotions towards politics play an important, but often overlooked, role in explaining vote choice.