Abstrakt:
Voter turnout in regular parliamentary or presidential elections is a very frequent topic, mainly in the recent debate about its decline. This article works with several theories explaining variations in voter turnout that are subsequently applied on referendums in Latin America. Referendums as one of the main pillars of direct democracy are not, in this regard, sufficiently scientifically explored topic. In this point of view, the region of Latin America is an ideal environment for research of the given subject, because there are many cases of direct democracy applications in the institutional configuration as well as in practice. Voter turnout in referendums is set into a comparative perspective using regression models that allow the researcher to monitor possible correlations and control explaining values. The analysis employs aggregate data from various statistical databases (e. g. V-Dem, Polity IV, World Bank) as well as secondary data from available comparative literature focused on elections (and referendums) in Latin America. The results indicate very high circumstantiality of referendums and the respective societies that apply them in politics. Some institutional settings have been found to positively influence the turnout, mainly compulsory elections.
