The Importance of Descriptive Representation & Cue-Taking in Free List Local Elections

Abstract:

With this study, we add to the literature on cue-taking and descriptive representation by focusing on a local, low-information context. Specifically, we empirically analyse voting patterns in the 2024 municipal elections in Mannheim, Germany. In this election, a free-list ballot design allowed voters to distribute 48 votes among 508 candidates of multiple parties, making it an ideal case to study these prominent theories of vote allocation. First, we show that party affiliation, ballot position, and incumbency significantly predict vote totals as expected. Ballot cues indicating candidates’ occupations and (to a limited extent) their gender additionally affect electoral outcomes. These findings complement previous experimental studies on ballot cues with observational research to better understand voter decisionmaking in complex real-world electoral settings. Using multilevel models, we further uncover a strong relevance of (geographical) representation and a home-district advantage in particular: Candidates receive about two-hundred per cent more votes in their own residential districts compared to their results in other districts. They also achieve substantially better results in districts with similar geographic location as their own. Lastly, voter age dynamics suggest that candidates benefit electorally when their age aligns with the demographic profile of the respective district.