Does district magnitude matter? The case of Taiwan

A sizable literature on electoral institutions argues that proportional electoral rules lead to higher voter turnout. However, recent work finds little evidence that the effect generalizes beyond western Europe and suggests that the theoretical arguments in the literature remain sparse, incomplete, and contradictory. I use a well-chosen data set to resolve the problem of omitted variable bias and Bayesian model averaging to address model uncertainty. I use Bayes factors to evaluate evidence both for and against the null hypotheses and find that the proportionality of electoral rules exerts no meaningful effect on turnout or any of the theoretical mechanisms I test.