Abstract
Voters are systematically unrepresentative of the eligible electorate. Many reforms intended to increase turnout and improve the representativeness of the voters have had underwhelming results. The ability to cast votes on a mobile device could potentially have more powerful effects since mobile voting would drastically lower the cost of voting, particular for certain underrepresented groups. In 2018, West Virginia became the first U.S. state to utilize mobile voting in a federal election, allowing it for overseas voters from 24 of its counties. I utilize this trial to assess the likely effects of mobile voting on the size and composition of the voting population. Implementing a differences-indifferences design with individual-level administrative data, I estimate that the ability to vote with a mobile device increased turnout by 3-5 percentage points, a large effect relative to other electoral reforms. At the same time, novel survey data shows that many Americans are understandably wary of online voting.