Why do some countries use mechanisms of direct democracy (MDDs) exceptionally frequently, others seldom, and still others never? No theory, so far, has comprehensively offered an answer to the previous question, and most conjectures are usually based on anecdotal or impressionistic evidence from a few selected cases. This chapter contributes to filling this lacuna by statistically analyzing the use of MDDs (either top-down or citizen-initiated) on an annual basis for every country worldwide.

Using a series of multivariate statistical analyses specially designed for “rare” events – such as negative binomial – and ordinary least squares (OLS), this study undermines many of the assumptions in the literature while confirming others. The findings suggest that both families of MDDs (citizen-initiated and top-down) are positively associated with the level of democracy, the age of the regime, the type of colonial heritage, and the number of MDDs used in neighboring countries in the immediate past. Contrary to commonly made theoretical assumptions, neither the size nor the social diversity of the population has any effect on the use of MDDs.

A great consensus exists on the contemporary challenges of representative democracies in several corners of the world. Some argue that a possible improvement for this state of affairs would be “to give back” to the citizens the capacity of control (reestablishing accountability), turning them into veto players through the activation of MDDs wherein specific policies could be proposed or vetoed and elected officials removed from their offices.