The Viability of Voting Advice Application Data for the Analysis of Coalition Formation

Modern coalition theories consider both the office and the policy motivation of political parties. Consequently, the application of these theories requires data on parties’ strengths as well as on their ideological positions. While the first – the seat distribution – is known as soon as the electoral result is announced, the estimation of party positions is a more complex and ambiguous task. Traditional methods such as expert surveys or analyses of party manifestos face various challenges. The question of how manifestos can be aggregated to party positions, for example, is highly disputed, and expert surveys simply are not available for every election. Recently, Voting Advice Applications (VAAs) have been provided for many elections. While VAAs have originally been developed to assess the congruence between parties and voters, they also enable the analysis of disagreements between parties. Due to the broad applicability and ease of collection of VAA data, it presents itself as a potentially efficient alternative to traditional data sources. However, there are also concerns if VAA data adequately reflects proximities between parties. Are the questions posed in VAAs crucial for parties during coalition negotiations? Do these questions delve deeply enough to identify substantial differences? Is it possible for highly specific questions to adequately reflect general ideological conflicts? With regard to the potential but also to the skepticism surrounding the use of VAA data for the analysis of coalition building, we contrast results of coalition formation models based on VAA data with those based on expert surveys and party manifestos. Our results show that, with few reservations, VAA data is suitable for estimating the policy component of coalition theories.