In comparative studies of turnout, compulsory voting is often considered as one of the independent variable related to the electoral system that contributes the most to the explanation of the levels of electoral participation. This paper attempts to control for this effect by analyzing turnout in a compulsory voting setting. Indeed, even if voting is compulsory, important differences in turnout can be observed. In Belgium and Ecuador, turnout has always been rather high and very stable over time, but not across districts. Using data from local elections in these countries for the last local elections of 2012 and 2014, this paper explores quantitatively which institutional, sociological and political factors explain differences in turnout, when the most important aspects of the electoral system are held constant (compulsory voting, PR, etc.).
