Electoral polarisation and voter turnout: Insights from the Dutch general elections

In his influential book, An Economic Theory of Democracy, Anthony Downs reaches a rather stirring conclusion, suggesting that abstention is the rational choice for most voters. Yet, voters show up in large numbers on election days, a puzzle which has been labelled as "the paradox of not voting". This has prompted numerous scholarly responses attempting to solve this puzzle and explain why people vote. While some suggest that the decision to vote or not is largely random, many have proposed solutions that address either the returns or the costs components of the Downsian utility function. Polarisation has been suggested as a solution to this puzzle that can address both the returns and the costs components. However, the scholarly discourse on the link between polarization and voter turnout is marked by two aspects which we address in this paper. First, early works studying this relationship focused mainly on the impact of ideological polarization on voter turnout, while recently more and more scholars are highlighting the impact of affective polarization on turnout. To the best of our knowledge, the relationship between turnout and electoral polarization remains unstudied. Although it is closely linked to ideological polarisation, we make the case that electoral polarisation is conceptually distinct enough and deserves special attention. Thereafter, we study its link with voter turnout. Second, most scholars seem to agree that, by turning elections into high-stake contests, polarisation piques the interest of larger numbers of voters in the electoral contest, driving up turnout. This perspective is notably optimistic, as it suggests that polarisation can remedy low rates of voter turnout, which has long been viewed as a democratic malaise. Empirically, it implies the presence of a positive correlation between polarisation and turnout. Examining the link between turnout and electoral polarisation, meanwhile, we present strong evidence in contrast with this optimistic view. In our study of electoral outcomes from the Dutch general elections in 2010, 2012, and 2017, we found a negative, substantial, and significant link between voter turnout and electoral polarization. We argue that as fewer voters show up on election days, voters of radical parties are overrepresented, driving up the level of electoral polarization. These findings emphasize the contribution of this paper, which addresses the gaps in literature highlighted earlier. Although our findings challenge the optimistic view described above, the conclusion does not have to be pessimistic, particularly regarding electoral polarization. Quite the contrary, our findings suggest that electoral polarization could be moderated if more voters participated on election days.