Which voting rule is most likely to choose the “best” candidate?

Abstract

One criterion for evaluating voting rules is the frequency with which they select the best candidate. Using a spatial model of voting that is capable of simulating data with the same statistical structure as data from actual elections, we simulate elections for which we can define the best candidate. We use these simulated data to investigate the frequencies with which 14 voting rules chose this candidate as their winner. We find that the Black rule tends to perform better than the other rules, especially in elections with few voters. The Bucklin rule, the plurality rule, and the anti-plurality rule tend to perform worse than the other 11 rules, especially in elections with many voters.