Examining voter turnout using multiscale geographically weighted regression: The case of Slovakia

Abstract

Voter turnout is an essential aspect of elections and often reflects the attitude of a country's population towards democracy and politics. Therefore, examining the distribution of voter turnout and determining the factors that influence whether or not people will vote is crucial. This study aims to find significant factors that underlie the different levels of electoral participation across regions in Slovakia during the 2020 parliamentary elections. In this interpretation, special attention is paid to the ability of the main theories of voter turnout to explain the behaviour of Slovak voters. The primary analytical tool is multiscale geographically weighted regression, which represents an advanced local regression modelling variant. The results indicate that the multiscale geographically weighted regression is superior to the global ordinary least square model in virtually all aspects. Voter turnout is generally higher in economically and socially prosperous localities and regions, which is in line with the societal modernisation theory. Additionally, factors connected to mobilisation theory and the concept of 'left behind places' also proved to be valuable. However, in other cases, such as with the share of retirees and potential habitual voting, the outcomes were not overly convincing, and further research is required.