Differences in early voting uptake in Sweden and New Zealand: comparing ‘election ecosystems’

Election administration reforms, such as extending access to early voting, are meant to increase turnout and political participation by lowering its costs. While there is research to suggest early voting does matter, some studies indicate the reverse; that early voting and other convenience voting reforms may magnify the existing socioeconomic bias in the electorate and facilitate participation for well-represented groups (Berinsky 2005, Karp and Banducci 2001, Kropf 2012). What then might account for variation in who votes early? Using data from the Swedish Election Authority, Statistics Sweden, New Zealand Electoral Commission and Statistics New Zealand I analyse and compare the patterns of aggregate-level socio-demographic and institutional correlates of the early voting rate in the 2014 national elections of both countries. Despite similarities in electoral institutions, such as proportional representation and the length of the early voting period, aggregate-level patterns suggest there are cross-national differences in who votes early. In New Zealand the early voting electorate and the overall electorate in New Zealand look similar except for a difference in median age. This contrasts with Sweden where the uptake of early voting at the kommun level shows negative correlation with median income. I hypothesise that procedural differences may help to explain this variation; for example, the implementation of voter registration procedures and the opening hours of early voting sites influence who will vote early. To date, these administration-level factors have seldom featured in comparative analysis of convenience voting. Thus, I argue that theorising the impact of reforms in electoral procedures should take into account interactions of the reform with the rest of the institutional environment, what Huefner et al. call the ‘election ecosystem’ (Huefner et al. 2007). As such, this paper makes a contribution to the theoretical debate on convenience voting which has tended to focus on majoritarian systems. Its conclusions also offer a word of caution to policymakers interested in using early voting to increase turnout or to reduce the socioeconomic bias of turnout.