New Trends in the Research on Voting Advice Applications

In Europe – as well as in a growing number of countries across the globe – Voting Advice Applications (VAAs) have been widely used for elections on different levels (subnational, national, transnational) and for elections of various actors (parliaments, presidents). VAAs are online tools designed to help voters comparing their policy preferences on major issues with the respective positions of political parties/candidates. As a result these tools indicate the issue proximity between the individual voters and parties/candidates running for election. In the last years, VAAs have become an established sub-field of social science research resulting in national projects, publications and networking as well as in the formation of an ECPR Research Network in 2015. Previously isolated VAA research efforts and projects have thus been pooled in an effort to establish a research community focusing on the idea, design and implications of these tools. By bringing together all European scholars working on VAAs, the Research Network has focused its attention on the wealth of research that is examining the effects of VAAs on parties, candidates, and voters, and how VAA design choices affect the voter advice given and the subsequent voter behavior. Recently, the study of VAAs has been significantly advanced especially in terms of the data VAA research draws on and the methodological designs that have been applied to understand the role of these tools in representative democracies. Moreover, there is an increasing number of studies using VAA data for addressing a broad spectrum of question not all of them directly related to the making of these tools. For instance, recent projects such as the Research Network's "VAA Party Positioning Project" have highlighted the potential in VAA-generated data for studying party positions over time and across countries, allowing for comparative analyses of the characteristics and development of parties and party systems. The section focuses on these new developments and is a follow-up of the successful VAA sections in Bordeaux (2013), in Glasgow (2014), in Montreal (2015) and in Prague (2016), each of which saw about 20 papers presented and 30 authors attending. This time, the focus of the individual panels will be to discuss new trends and tendencies in VAA research overcoming the limitations of traditional theoretical and methodical approaches of VAA studies. The panels of the section provide a platform to exchange the findings of VAA research with research from other fields of political science. The Section is endorsed by the ECPR Research Network on Voting Advice Applications. Planned Panels: (If you are interested in submitting a paper for one of the four panels, please contact the respective panel chair directly.) Panel 1: The Design of Voting Advice Applications: Theoretical Background and Methodological Challenges Chair: Kostas Gemenis (k.gemenis@utwente.nl) Even though Voting Advice Applications have been around for more than 25 years, the stability and consistency in their design is particularly surprising. VAAs designed today often look nearly identical to VAAs designed more than a decade ago. The purpose of this panel is to evaluate and challenge some of the established practices, and explore the implications for the quality of recommendations generated by VAAs. The panel therefore invites papers that focus, among other things, on the various methods used in VAAs to a) select, phrase, and frame questions, b) estimate the positions of parties and candidates, c) match users to parties or candidates, and d) visualize the results. Papers that create links between the methods used in VAAs and other areas of research, presentations of innovations in terms of design, and studies that compare the different methods based on randomized experiments are particularly welcome. Panel 2: The Effects of Voting Advice Applications: VAA Influence on Individuals, Organizations and Systems Chair: Martin Rosema (m.rosema@utwente.nl) The study of VAAs has developed into a fascinating area of research in political science, linking up nicely to several of its subfields. One of the topics that scholars paid attention to is the effects that these tools have on voters. VAAs have often been promoted as tools that are capable of increasing the political knowledge of citizens and boost electoral turnout. Several studies have provided empirical support for the hypothesized effects, although the estimates of their size vary substantially. With respect to candidate and party choice, evidence concerning the influence of VAAs has also accumulated and researchers have identified several factors that moderate these effects. Having established some basic insights, the time seems ripe to deepen our understanding of the influence that VAAs have. In this panel we therefore welcome papers that further develop our understanding of the extent and the conditions under which voters are being influenced by the use of VAAs. Furthermore, the panel also welcomes papers that focus on other effects that VAAs may have. In particular, the question arises if the immense popularity of VAAs in several countries has affected the behaviour of political parties. For example, does the widespread use of VAAs stimulate political parties to take positions on issues more strategically, taking into account the views of the electorate? In some countries interest groups have developed their own VAAs, so one may wonder how the role of such organizations in election campaigns has been affected by the rise of VAAs. Finally, the panel also welcomes papers that adopt a broader focus and analyse the impact of VAAs on the political system as a whole. More specifically, does the use of VAAs increase the quality of representation and hence improve the functioning of the political system? Researchers who explore such questions – no matter which methodology they use or what countries they focus on – are warmly invited to send their paper proposals for this panel. Panel 3: Voting Advice Applications, Party Positioning and Party Systems: Measuring the Dimensionality of the Political Space across Countries and Time Chairs: Eric Linhart / Ioannis Andreadis (eric.linhart@phil.tu-chemnitz.de) The estimation of party positions in policy spaces is an important task for numerous political science subfields. It is needed, e.g., for analyses of electoral campaigning, voting behavior, coalition building, legislative processes, and many more. Voting Advice Applications (VAA) are a tool originally designed for another purpose – namely giving voters an orientation about how strongly they agree or disagree with programs of various parties – but have become an interesting alternative and/or supplement to so far existing data bases like expert surveys or party manifestos. However, the estimation of party positions is a highly sensitive challenge, and we still lack knowledge about the empirical connections of and differences between the various methods. This panel aims to bring together authors trying to reduce this gap on the basis of VAA analyses. Papers in this panel can include (but are not limited to) - the development and application of VAA based methods to estimate party positions, - the comparison of party positions based on VAA data with those generated by other methods, - analyses of party competition, voting behavior, coalition building or similar on the basis of VAA data, or - VAA based methods to estimate the dimensionality of policy spaces. Both theoretical and empirical papers are welcome. Panel 4: Voting Advice Applications and Social Computing: VAAs as Driving Force and Mirror of Social Change Chair: Vicky Triga (vasiliki.triga@cut.ac.cy) The panel aims to attract interdisciplinary papers on social computing aspects of VAAs. VAAs can generate new kinds of data on an unprecedented scale to reveal patterns of individual and group behaviour. This much has already been established by the nascent VAA literature. However, apart from a few notable examples the field has been largely dominated by electoral studies approaches. While this has been a wholly positive development, the frontiers of what can be done remain relatively unexplored, especially from more interdisciplinary perspectives. To name but a few frontier domains, further developments involving massive online randomised experimentation, social network analysis, machine learning and the application of data mining techniques are ripe for exploitation. Such analyses may provide insights on broader questions of social change, such as the promotion of political participation, the use of social vote recommendation processes, the implementation of automated ‘unbiased’ methods of selecting policy issues as well as estimating parties’ positions on the respective issues, etc. To this end, the aim of this panel is to bring together social scientists with an interest in computer science and socially literate computer scientists to explore how the interdisciplinary methodologies and techniques of computational social science can be applied to VAA generated data on the one hand, and how this may contribute to social change on the other hand.