E pluribus unum:

Äänestysmenettelyn valinnan ongelmia

Authors

  • Hannu Nurmi

Abstract

Thirteen voting procedures are compared with respect to nine criteria. The binary procedures discussed are the parliamentary voting procedure (or pairwise comparison with the simple majority rule), Copeland’s and Schwartz’ methods as well as the maximin method. The one-stage methods dealt with are: the plurality principle, Borda count and approval voting. The following multistage non-binary methods are investigated: Black’s, Nanson’s, Hare’s and Coombs’ methods along with the plurality runoff procedure. The criteria of comparison are: Condorcet’s two criteria (winner and loser), monotonicity, Pareto, weak axiom of revealed preference, path independence, consistency as well as two implementation criteria: simplicity and easiness. It transpires that the performance of the most commonly used procedures — the parliamentary voting procedure, the plurality principle and the plurality runoff — is singularly unimpressive: if we assume that the voters are capable of producing a strict ordering of the alternatives, each of these methods is dominated by at least one of the other procedures. Not all are dominated by the same procedure, though. On the other hand, the approval voting fares rather well on all counts except on Condorcet’s criteria.

Downloads

Download data is not yet available.
Section
Articles

Published

1981-09-01

How to Cite

Nurmi, H. (1981). E pluribus unum:: Äänestysmenettelyn valinnan ongelmia. Politiikka, 23(4), 410–437. Retrieved from https://journal.fi/politiikka/article/view/150413