How Incumbency Shapes Electoral Competitiveness in Ranked-Choice Voting

We provide the first systematic analysis of electoral competitiveness in American ranked-choice voting (RCV) elections. First, we develop a novel dataset of the margin of victory in American RCV contests from 2004 to 2024 by integrating a state-of-the-art optimization algorithm with over 38.9 million candidate rankings submitted by actual voters. Second, we propose a theoretical explanation for how the presence of an incumbent may shape electoral competitiveness under RCV. We show that incumbent-held seats are, on average, less competitive and have fewer candidates than open races. Using regression discontinuity with non-ignorable sample selection, we also illustrate how incumbency may allow candidates to cultivate favorable rankings under RCV. Finally, we also show that incumbent-held seats are, on average, less likely to have second and higher rounds than open races, making voters' lower preferences count less. These findings have broad implications for election reforms, the incumbency advantage, and electoral competition.