Should voters be required to rank candidates in an election?
Abstract
We compare the Condorcet Efficiencies of the plurality rule, the negative plurality rule, and the Borda rule to address the question of what might be gained by using a voting rule that requires candidate rankings. Unlike previous analyses, we consider only those voting situations for which the three rules determine different candidates as winners, because these are the cases where the Condorcet Efficiencies might actually differ across the three rules. After assessing the theoretical as well as the empirical Condorcet Efficiencies, we find that, despite considerable differences between the properties of the theoretical framework and the characteristics of three sets of empirical ranking data, all four analyses suggest that there is a considerable benefit in asking voters to submit candidate rankings.