ESRA 2019 Programme at a Glance
Public Opinion and Electoral Politics in an Era of Political Discontent 2
|Session Organisers|| Dr Roula Nezi (GESIS-Leibniz Institute for the Social Sciences)
Dr Theofanis Exadaktylos (University of Surrey)
|Time||Friday 19th July, 09:00 - 10:30|
Over the past nine years the European Union has faced a series of social and political challenges that affected citizens’ political behaviour, political attitudes, and party systems. The European Union in particular, but also non-EU member states, confronted a series of events such as the economic crisis and the refugee crisis, which, coupled with the prominence of austerity politics, have given rise to unpredicted political and electoral outcomes such as the rise of populist parties - both from the right and the left of the ideological spectrum - the rise of authoritarian politics, the rise of political forces questioning the future of European integration as well as the overhaul of traditional parties.
These phenomena give rise to important questions for scholars working in the area of public opinion and elections. Can the existing theories of electoral choice explain the surprising electoral outcomes witnessed in many countries? What is the role of emotions in political behaviour? Are the recent electoral shocks a result of a crisis of confidence and trust facing mainstream political parties and the rising disconnect of citizens? Is the growing support for populist parties rooted in austerity politics or is based on changes in peoples’ values and emotions?
This panel welcomes papers on a wide range of topics related to public opinion, elections, voting behaviour, and election forecasting such as:
voter turn out
political participation and pathways to engagement
vote for populist parties
authoritarian attitudes and values
emotions and appraisals
We welcome papers using single case studies but we especially encourage comparative/longitudinal studies. Proposals should encourage the conference’s main theme “survey research in the changing data environment”. We also welcome papers that propose new theoretical approaches in the study of public opinion and elections and are empirically or methodologically innovative.
Keywords: public opinion, electoral politics, political participation, values, survey research
Impact of Unequal Turnout on European Election Results – Evaluation of Traditional and New Applications
Mr Stefan Haussner (University of Duisburg-Essen) - Presenting Author
How would election results in the European Union change, if everybody showed up to vote? Despite being counterfactual, this question nevertheless has important policy implications and was subject of many research articles in the field of electoral studies – without a clear result. While conventional wisdom among media and politics suggests that left leaning parties would benefit from higher turnout rates, empirical political science often failed to confirm this hypothesis. The vast majority of research confirms a strong correlation between socioeconomic status and electoral participation, showing less participation among the underprivileged. Similarly, the majority of studies hint – not prove – that socioeconomically underprivileged groups traditionally lean more towards left-wing parties. However, only a minority of studies acknowledge major changes of election results with universal turnout. A multitude of academic attempts used regression/imputation approaches to predict the party preferences of nonvoters.
The scope of the proposed paper is two-folded: First, it will evaluate existing approaches of empirical political science and discusses why previous attempts failed to confirm changes in election results. Second, using the last rounds of ESS data, the paper introduces and assesses the application of Random Forests as a potential enhancement. Algorithms from the field of machine learning are often more flexible than traditional regression procedures, but therefore less interpretable. Since the question whether election results would change given universal turnout is more a question of classification rather than interpretation, this – at least for political science – new algorithm might be a promising new approach to answer the question of impact of unequal turnout on election results.
The project is placed in a comparative European Union context, covering a time period of crucial transitions of party systems during the economic crisis. The proposed paper will compare the quality of concrete applications of Multiple Imputation and Random Forests, which is interesting academically and could have a policy implications, as well.
Eroding Trust: Regional Effects of Consecutive Crises
Mrs Ann-Kathrin Reinl (GESIS) - Presenting Author
Dr Christina Eder (GESIS)
Professor Alexia Katsanidou (GESIS)
Interpersonal trust rates declined in many states during the European Financial and Migration Crises. Former research conducted on political movements in times of crises revealed a connection between strong populist groups and low levels of political trust (Algan et al. 2017; Fieschi & Heywood 2004). Therefore, research on the development of trust is of high relevance considering the current rise of populist parties all over Europe. By studying political trust as a dependent construct, we want to shed light on the question if the crises hitting the European Union during the last decade had a sustainable effect on public trust rates. To that end, our statistical analyses focus on Germany. During the past years, many comparative studies on the development of various forms of trust have been published (Foster & Frieden 2017; Van Erkel & van der Meer 2016). However, no study to date analyzed the case of Germany in-depth. We think that the German case is of special interest for various reasons. Firstly, some regions were more hit by economic downturn during the Euro Crisis than others. Secondly, regions located on the border to Austria experienced particularly high influx rates of asylum seekers during the Migration Crisis in 2015. We assume that low levels of trust in national and European institutions correlate with the level of crises consternation. By employing regional data of the European Social Survey from 2008 to 2016 we are able to obtain information on the development of trust rates in Germany, otherwise lost in the national level. To answer our research question, multilevel logistic regression analyses are conducted. Moreover, the impact of political events on trust rates is measured via a quasi- experimental design. Our statistical models control for the effect of political events within the time period of the Euro and the Migration Crises.
From One Crisis to Another: How Transnational Conflict Lines Shape in the European Parliament
Professor Alexia Katsanidou (GESIS - Leibniz Institute for the Social Sciences) - Presenting Author
Professor Zoe Lefkofridi (University of Salzburg)
How do changes in the dimensionality of the political space impact the quality of representation in the EP? We pursue this question in two stages: first, we examine which issues matter most for voters’ choices of national parties in European elections; and second, we analyze which issues drive the sorting of national parties in EPGs. More specifically, we observe the role of outliers overtime: do national member parties with positions deviant from the median “assimilate” or are they the trend-setters for the future? We focus on two European Parliament elections (2009 and 2014) and utilize data on political party positions (EUProfiler and EUandI) and on voter positions (European Election Study); in detail, we examine three issue dimensions: economic left-right, immigration and EU integration. Our findings will address the wider question on whether a reconfiguration of traditional cleavage lines is taking place across Europe thus redefining representation in the EP.
Using a Machine Learning Voting Advice Application to Capture the Distance between Parties and Voters in the European Political Landscape.
Mr Javier Padilla (London School of Economics) - Presenting Author
Mr Enrique Chueca (King College London)
Mr Guillermo Romero (University of Southamptom)
Voting Advice Applications (VAAs) using machine learning techniques can be used to capture the conceptual political space of the users and explore the ideological distances between voters and political parties in high and low dimensional political landscapes. In this paper, we use the data from two VAAs launched for fostering turnout and political knowledge at the European Parliamentary Election: the EU-Vox 2014 and the EU-Vox 2019. Thanks to machine learning techniques, we can measure accurately whether in this five years of economic and democratic crisis (raise of left and right populism in several European countries, secessionist movements in Spain and general political discontent among most European countries) there has been in the VAA users a 1) change in the saliency of the policy issues in the European countries; 2) changes in the saliency of the main three dimensions at the European Parliamentary Election (Economic, Social and European dimension); 3) changes in the ideological distances between users and political parties; 4) changes in the way of comparing policy issues between users and voters. To do this research, we have the database of the EU-Vox 2014 (more than 3 million total users) and the EU-Vox 2019 (which will be launched for the next European election, and it is expecting to have a higher repercussion than the previous one). We plan to do a cross-national and cross-age study on the changes on the European political landscape of the users, addressing whether populist parties are covering political spaces which were unoccupied by mainstream traditional parties.