Egalitarianism vs. Utilitarianism in preferential voting

Abstract

Democratic societies base much of their decisions on voting procedures that involve aggregation of individual votes into a winning solution. While for two candidates majority voting can provide satisfactory results, for three or more candidates the winner depends on the voting method employed. In this chapter we analyse preferential voting, where voting ballots consist of a ranking of candidates. We first study the classical Condorcet criterium introduced to maximise the total satisfaction of voters, i.e. the utilitarian criterion. We then complement it with a recently introduced method to minimise the total un-evenness of the rewards, i.e. the egalitarian dimension. We show, through targeted examples and analysis of synthetic vote data, that the new dimension may lead to more fair results, and can provide resilience to radical voter opinions.