For whom do local and regional elections matter? Localized versus second-order voting in municipal and county elections in Norway

Paper proposal for consideration of panel 1 on ‘Voters’ Behaviour in Sub-National Politics’. Recent election survey research has revealed significant levels of party switching between national and subnational elections. Typically, between a quarter to one third of the voters indicate that they vote for a different party in subnational elections than their nationally preferred party (Berg and Oscarsson 2020; Elklit and Kjær 2009). This has inspired a literature on explaining vote choice in subnational elections which pre-dominantly uses the second-order election model as a theoretical starting point (Gendźwiłł, Kjaer, and Steyvers 2022). According to this model, voters base their subnational vote choice on the governmental status of parties in national government. Voters who are discontent with the performance of parties in national government vote for new, small, or local parties or parties in national opposition in subnational elections. In this view, subnational elections are nothing more than ‘barometers’ or ‘referenda’ for the popularity of parties in national government. Recent research reveals that a second-order election view has limited explanatory value, to the very least, the model does not provide for a comprehensive understanding of the causes underlying the subnational vote (Lineira 2016; Kjær and Steyvers 2019). Furthermore, most research on subnational voting exploits aggregate data simply because individual-level survey data tend to be quite rare for subnational elections, especially for unitary countries. This paper takes up the question under what conditions voters tend to base their vote choice on subnational issues (localized voting) rather than on national politics (second-order voting). The theoretical innovation of this paper is to develop a model of localized voting that conceptualizes the conditions under which voters tend to hold subnational politicians rather than national politicians electorally accountable. The empirical contribution of this paper is that its analysis draws on highly innovative survey questions that asked voters to evaluate the importance of a range of issues while casting their vote. In addition, another set of survey questions tapped basic vote motivations and asked respondents to indicate the extent to which they used their vote prospectively or retrospectively in relation to subnational and national government. These survey questions were fielded in the Norwegian Citizen Panel wave 19 which took place one month after the municipal and county elections of September 2019. The sets of innovative survey questions and the comparison between municipal and county vote choice enable the identification of the conditions under which voters are likely to cast a localized or a second-order vote in subnational elections.