Adopting compulsory voting (CV) legislation is expected to produce near-universal turnout, which in turn is assumed to iron out class-based differences in political influence and representation. The article traces the historical process generating the sequential adoption of CV in 8 of the 17 Cisleithanian crownlands of the Austro-Hungarian empire in 1907 and 1911, and leverages a difference-in-differences (DiD) method to estimate its causal effect on turnout and voting patterns in elections to the Imperial Council. Exploiting unique data on the turnout of citizens based on their occupational categories, it further examines whether the adoption of CV attenuated class bias in turnout. Despite a large boost to turnout, CV neither increased support for parties representing the working classes, nor attenuated the class bias in turnout.

It has normally been argued that because compulsory voting systems present higher turnout rates relative to voluntary voting systems, they do not generate as many biases between different groups of voters. This article qualifies that view. It argues that in cases in which compulsory voting does not ensure near-universal participation, there is no certainty that switching to voluntary voting will increase inequalities. This issue is examined by looking at Chile, a democracy that moved from compulsory voting to voluntary voting in 2012. The research finds that while the reform generated class bias in urban districts, it also substantially reduced age bias and, in national elections, equalized participation between small and large districts. The conclusion is that abandoning compulsory voting does not necessarily increase turnout biases, since much depends on the structure of preexisting biases and how these are conditioned by particular electoral institutions.