Electoral Competitiveness and Turnout: How System and Preference Uncertainty Impact the Decision to Vote

Abstract

The relationship between electoral competitiveness and turnout forms the foundations of understanding voter behaviour. The more competitive an election, the more ballots that should be cast. This simple and longstanding association means that some form of competitiveness is now "used as an explanatory variable as a matter of course" (Denver, Hands and MacAllister, 2003:174) in the study of elections. Yet, the relationship was established in stable, two-party, majoritarian systems and focuses only on the aggregate level. Moreover, turnout is becoming increasingly difficult to predict (House of Lords, 2018; Biddle 2019) which suggests a weakening of the association. Instead, I argue that changes in the party system have altered the competitive environment, which in turn requires the revisiting of competitiveness itself. This research establishes two new conceptual terms for electoral political science. The first is system uncertainty, which represents the competitiveness of the district in which a contest is being held. The second is preference uncertainty, which is experienced at the individual level when a voter does not prefer one party over all others. Additionally, two new measurements for these concepts are constructed. The effects of these types of competitiveness on turnout are tested using British Election Study longitudinal data. The findings show that system uncertainty fosters turnout, in line with previous literature. In contrast, there is evidence that preference uncertainty fosters abstention and that the competitiveness-turnout relationship is moderated by individuals' risk aversion. This has widespread implications for electoral behaviour literatures, and could explain why many people vote when it is irrational to do so.