The Turnout Paradox: Why Changing Electoral Experiences Trump Changing Social Characteristics in Driving Voter Turnout in Advanced Democracies

This paper challenges the dominant paradigm of turnout studies by proposing that turnout variations occur because elections differ, not because countries differ; and certainly not because people differ. Indeed, we take the radical position that what matters if we want to understand overtime and cross-national turnout variations is differences in the ways in which citizens are affected by the electoral experience at different times and in different countries. Characteristics of individuals, including characteristics that affect their responsiveness to the electoral situation and to the mobilizing efforts of parties and other organizations, determine who will vote and who will not (except in rare instances of effectively perfect turnout), but equivalent characteristics at the aggregate level cannot be expected to have similar effects. This paper sets out the theoretical basis for rejecting the dominant paradigm, and develops a parsimonious model that explains turnout variations since 1945 in a universe consisting of 21 out of the 22 advanced democracies that have held elections continuously since World War II. The model explains more than 90 percent of variance in turnout across the 354 elections concerned. More importantly, it tracks with considerable accuracy the variations in turnout that have occurred since 1945 in virtually all of the 21 individual countries. The model explains the slight decline in turnout that has recently occurred in advanced democracies – a decline that appears to be due to a contemporary reduction in the competitiveness of elections in many countries that has had the effect of reducing their importance to voters. 2 In recent years it has become commonplace to note declining turnout at elections worldwide and books and articles with titles like The Disappearing American Voter (Teixeira, 1992) or "Exploring Declining Turnout in Western European Elections." (Flickinger and Studlar, 1992) are almost commonplace. A heavily funded project at Harvard University's Kennedy School of Government during the run-up to the 2000 presidential election was entitled "The Vanishing Voter Project." In Britain, the government authorized experimental changes in the voting procedures at local elections to find ways to increase future turnout, in Ireland a minister tried placing each candidate's picture on the ballot paper to see if that would help, while in Sweden senior officials considered the drastic possibility of doing away altogether with elections to the European Parliament since turnout at these elections is so low. There certainly appears to be some truth in the generalization that turnout is declining …