Turnout and Competitiveness in Space and Time.

Changes in the intensity of party competition drive changes in voter turnout, especially among young voters. The basic proposition has been established for elections in general (Franklin 2004) and for Canadian elections in particular (Johnston, Matthews and Bittner 2005). The details of the relationship, however, remain murky. Two related questions stand out. First, what mechanism links electoral competitiveness to the individual voter's turnout propensity? Strategic mobilization of supporters by parties and candidates seems a likely suspect (Rosenstone and Hansen 1993), but there is also evidence that voters take cues from competitive conditions directly (Cox and Munger 1989). Second, what rendering of electoral competitiveness in spatial and temporal terms is most relevant to variation in turnout? Previous work on Canadian elections has relied exclusively on contemporaneous measures of riding-level turnout (Endersby, Galatas and Rackaway 2002; Johnston, Matthews and Bittner 2005). What happens if we introduce measures aggregated to the province or nation and consider lags of electoral competitiveness?This paper addresses these questions through investigation of the spatial and temporal dimensions of competitiveness effects on turnout. The data for the analysis combines five waves of the Canadian Election Study (1988 to 2004) with indicators of electoral competitiveness measured at multiple geographical levels (riding, province, nation) and time points (current election, previous election). This multidimensional treatment not only speaks to the when and where of competitiveness, but also permits inferences about the mechanism(s) of the variable's impact on turnout. Current riding-level effects, for instance, imply an important role for the organizational efforts of parties, as an individual-level account of the relationship entails heroic assumptions about voter information. Lagged province-level effects, by contrast,are quite compatible with the low-information psychology of the typical voter. The analysis has payoffs both theoretical and practical. On the theoretical side, the paper promises to help sort out alternative accounts of the relationship between turnout and competitiveness, at the same time as it pushes the conceptual boundaries of competitiveness itself. On the practical side, the paper continues to illuminate the story of Canadian turnout decline, including treatment of the 2004 election. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]