Abstract
We introduce and study two new statistical models of ballot truncation – the process wherein voters neglect to rank every candidate during ranked choice voting (RCV). These models allow the incorporation of ballot truncation into simulations of ranked choice elections, resulting in more realistic predictions of likely outcomes. Our Independent Truncation model is based on a survival analysis approach, while our Plackett-Luce Truncation model extends the classical Plackett-Luce ranking model. Using ballot data from Cambridge MA elections, we motivate our new models and fit them to obtain parameter estimates. We then demonstrate our models in a simulation study to measure the effect of ballot truncation on minority representation in a hypothetical multi-winner election.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call