文摘
We study consensus measures that quantify the cohesiveness of the information generated when a group of decision-makers express their evaluations of a number of issues. Particularly, in social choice an approval consensus measure (ACM) is used to evaluate the degree of cohesiveness in a group of agents that have dichotomous opinions on the issues. In this paper we propose three novel consensus indexes that take advantage of the specific information about such opinions: namely, the Herfindahl–Hirschman ACM, the majoritarian ACM and the weighted majoritarian ACM. To illustrate their performance we apply them to the analysis of popular votes in Switzerland and Italy. The first analysis has a fixed population of agents (the cantons) and all votes are known. In the second analysis we have a variable population (the voters) and unknown individual votes. In both real case studies, we show empirical evidence that the new indexes can be used to assess consensus.