Changes that cause changes

Abstract

Beyond determining whether procedures can be manipulated, the real goal for any analysis of “strategic behavior” is to identify all settings where and when this can be done, who can do it, and what they should do. By applying the geometric approach of Saari [7, 8] to the Kemeny's Rule (KR), we demonstrate how surprisingly simple this analysis can be, we identify all three candidate KR strategic behavior, and we show how an almost identical analysis answers most other multiple profile concerns (e.g., the abstention paradox and when voters just make errors). We also introduce new measures, which can be used with any procedure, to compare strategic and other behavior involving “changes.” These measures help to identify settings where it may be more important to worry about honest mistakes than strategic voting.