Do people make their votes count? An experimental study on voter errors and electoral systems

We examine whether new and alternative electoral systems, each with unique ballot-marking characteristics, introduce administrative challenges that increase the likelihood of voter error. Using a within-subject survey experiment in the Czech Republic with a representative sample and a dual data-collection method (online and offline), we identify two key contributors to voting errors: electoral system complexity and the technical possibility of casting mismarked or invalid ballots. We propose theoretical dimensions of electoral system complexity—input structure, instruction length, and type of action—and provide a framework for studying voter behaviour and ballot-design effects. Based on these effects, we confirm that there is no direct trade-off between voting errors and precise preference expression. Online voting shows increased error rates compared to pen-and-paper voting. Our findings have practical implications for electoral reforms, showing that a full-ranking system and a conditional method pose challenges in terms of voter errors.