Smartwielen: Use and impact at the 2023 national and 2024 European elections

Well-functioning democracies need informed citizens. This is especially the case as voters have been shown to be generally not competent enough to cast a vote for the party that best represents their interests (Campbell et al. 1960). Voting Advice Applications (VAAs) such as the smartwielen tool implemented in Luxembourg since 2009 provide systematic comparative assessments of parties’ and candidates’ preferences over a range of policy issues on a single online platform; as such they reduce the costs to voters associated with the gathering and processing of electoral information (Pianzola et al. 2019). These digital platforms are designed to increase informed and critical public participation in the electoral process. They have become a vital part of online election campaigns worldwide, being used by dozens of millions of voters (Garzia and Marschall 2019).

This project will engage with a number of questions that have so far not been addressed, or only with a research design that did not allow to come to robust conclusions, in VAA research: first, how do voters actually use VAAs; second, what is its impact on users? We will in addition take advantage of the proximity of two elections in Luxembourg (2023 national elections and 2024 European elections) and the deployment of the smartwielenonline tool by a consortium consisting in the Université du Luxembourg, the Zentrum fir politesch Bildung foundation and the LISER – where the seconded researcher will be hosted – to address the potential differential impact of VAAs across first order and second-order elections.

The originality of the proposal lies in its combination of an existing outreach project and innovative scientific research orientations destined to trigger both a large social and a high scholarly impact. It encompasses all three types of VAA research – design, impact and use of VAA-generated data – identified by Dumont and Kies (2015). First, it will provide the most ambitious experimental pre-online phase of VAA design work ever envisaged through two waves of laboratory experiments aimed at learning about how voters actually use the tool, how they allocate their attention to different kinds of information and policy issues, and they react to different kinds of electoral information stimuli. Second, it will study the effects of VAA use with a cutting-edge research design involving a randomized field experiment between a pre-electoral and a post-electoral panel survey. An encouragement to use the smartwielen tool will be sent to a treatment group (half of panel members) before the election. Tracking these subjects' use of the tool will then allow to test hypotheses regarding VAA use and personalized advice on their attitudes and behaviour. In addition to overall effects, we will pay additional attention to first-time voters (who recently gained voting rights due to age or nationality acquisition) who are the most likely to benefit and therefore to be impacted by VAA use.

The unique data collection enabled by VAA research will further provide a wealth of new evidence to engage long-standing questions about political representation, inter- and intra-party competition, electoral campaigns, and voting behaviour in Luxembourg. Overall, the research stay will address gaps in the study of society and politics related not only to voters' electoral information environment but also, by extension, rising concerns about decreasing trust in institutions which can be affected by the discrepancy between the actual distribution of policy preferences in the voting population and its representation in political assemblies. It will ultimately contribute to this larger question by assessing the potential of VAAs to reengaging voters with the political process and policy-based politics VAAs may eventually contribute to rebuild trust between parties, their representatives or candidates, and the public