Elite polarization and voting turnout in Latin America, 1993–2010

Sergio Béjara*, Juan A. Moraesb & Santiago López-Caribonica Department of Political Science, San Jose State University, San Jose, CA, USAb Facultad de Ciencias Sociales, Universidad de la República, Montevideo, Uruguayc Departamento de Ciencias Sociales y Políticas, Universidad Católica de Uruguay, Montevideo, Uruguay CONTACT Sergio Béjar sergio.bejar@sjsu.edu Department of Political Science, San Jose State University, San Jose, CA, USASupplemental data for this article can be accessed at https://doi.org/10.1080/17457289.2018.1545775.ABSTRACTDespite substantial advances in political participation across Latin America, relevant data from the region reveals an important cross-national and temporal variation in voter turnout. Unlike existing region-specific explanations of voter turnout, but in line with research on advanced democracies, we show here that actual electoral mobilization in Latin America is linked to the ideological and programmatic shifts experienced by political elites since the 1990s. Specifically, we first suggest that by setting clear programmatic stances and differentiation polarization allows voters to better understand the costs and benefits of turning out to vote. When more public policy is at stake, elections become more consequential, and thus voters are more likely to turnout. Second, polarization facilitates the formation of party brands that structure the policy preferences of voters, easing their mobilization to the polls. Our results, using elite survey data and electoral turnout for eighteen Latin American countries during the 1993–2010 period, provide strong statistical support for the elite polarization-turnout hypothesis and remain robust to different specifications of the model. Substantially, an increase in one standard deviation in elite polarization increases turnout – measured as a percentage of registered voters – by about 7 percentage points. This effect increases to 9 percentage points when turnout is measured as a percentage of the voting age population.