Several countries have started to experiment with Internet voting (`i-voting'), often with the goal of increasing electoral participation. This paper considers another benefit of i-voting that has received less attention in the existing literature: the availability of software fixes that can help voters to avoid common mistakes that lead to their votes remaining uncounted. Independently of a possible effect on turnout, i-voting is thus likely to increase the effective turnout in terms of valid votes cast. I provide empirical evidence for this claim based on a case study of extended i-voting trials in Geneva canton, Switzerland. Using difference-in-differences estimation, I find that i-voting reduced the residual vote share by around 0.3 percentage points in referendum ballots. The effect increases to almost 0.5 percentage points in down-ballot measures. Such increases in effective participation could make a difference in closely fought electoral contests. However, in the case under study I find no evidence for a systematic shift in electoral outcomes.
