Stimulating Deliberative Democracy Through VAAs: the Addition of Argument Summaries and Fact-Check Labels

"During election campaigns, voters face the challenging task of deciding which party to support. This task is particularly demanding in multi-party democracies such as the Netherlands, where in the last national elections no less than 27 parties competed for the voter’s favor. To get some idea of what these parties stand for, many voters rely on Voting Advice Applications (VAAs). In these tools, users respond to a series of political attitude statements and receive a ranked list of parties that best match their preferences. Previous research has shown that VAAs can increase (perceived) political knowledge and voter turnout (see Munzert & Ramirez-Ruiz 2021 for a reflection). At the same time, empirical evidence indicates that voters often struggle to fully comprehend the attitude statements and make limited effort to find additional information outside the tool (Kamoen & Holleman, 2017). Moreover, from a deliberative democratic perspective, VAAs have been criticized for primarily facilitating issue voting by reducing political choice to the aggregation of individual policy positions, rather than supporting voters in engaging with the arguments (Fossen & Anderson, 2014). As a result, current VAAs may help voters identify preference congruence with parties but do little to stimulate reflection on why particular positions are held. In a series of two studies, we therefore examine the effects of incorporating argument summaries into VAAs on a range of outcome measures. In Study 1, both theoretically and practically educated voters (N =233) were assigned to one of three conditions: a standard VAA without supplementary information; a VAA+ including definitions of key terms, status quo information, and party positions; and a VAA++ containing all VAA+ information plus summaries of pro and con arguments. Initial results show that all types of supplementary information were used frequently, with argument summaries receiving particularly many clicks (M = 2.74 for practically educated voters; M = 7.74 for theoretically educated voters). In terms of evaluative outcomes, both the VAA+ and VAA++ were rated more positively than the standard VAA on a range of measures, while differences between the VAA++ and VAA+ emerged for the construct “opinion formation”. In Study 2, we compared two versions of the VAA++: one with a list of pro and con arguments, and one in which these same arguments lists were accessible, but the arguments were now fact-checked and accompanied by fact-check labels (“true”, “not true”, or “not checkable”). In an experimental between-subjects design, no differences between these conditions were found on a range of outcome measures. However, in a subsequent forced-choice task in which participants were presented with both versions, a large effect emerged in favour of the fact-checked condition. In the presentation, we reflect on the findings of both studies and discuss their implications for deliberative democratic theory, as well as for the normative and practical design of VAAs as tools for civic education."