Interventionism of voters: district size, level of government, and the use of preference votes

Abstract

This paper examines the use of preference votes under the open-list proportional representation system in the elections of assemblies at different levels of government. Our empirical analysis focuses on the elections held in Poland, where similar system is applied in elections of councils in three subnational tiers. This setting allows us to test the hypotheses concerning the impact of party magnitude and district size on the usage of preference voting. Earlier research demonstrated that the distribution of preference votes is heavily influenced by candidates’ ballot positions and their personal vote-earning attributes. While the ballot position serves as a cue for less-informed voters in all tiers, we demonstrate that the elections held in smaller constituencies, where voters are more proximate to their representatives, are more personal. This is reflected by the higher chances of changing the candidate order by using preference votes in constituencies characterized by the lower voters per seat ratio. We also find that preference voting matters more when party magnitude is larger. Our theoretical expectations are tested using logistic regression models, accounting for candidate- and list-level effects.