Refining measures of group mutual coherence
Abstract
The Condorcet efficiencies of plurality rule (PR), negative plurality rule (NPR), Borda rule (BR), plurality elimination rule (PER) and negative plurality elimination rule (NPER) were evaluated over parameters associated with six models of group mutual coherence in Gehrlein and Lepelley (Voting paradoxes and group coherence: the Condorcet efficiency of voting rules, 2010) It was found that BR was not always the most efficient voting rule, but it always performed quite well; while each of the other voting rules had identifiable regions of parameters in which they performed very poorly. By refining these parameters so that attention is focused on the particular model of group coherence that most closely reflects the voters’ preferences in a given voting situation, these conclusions are modified. The comparison of BR to PER and NPER changes significantly. The comparison of BR to PR and NPR remains similar, but the differences in the relative comparisons of efficiencies are somewhat reduced.