The Limits of Vulnerability to Manipulation as an Argument for Electoral Reform

Abstract

The field of social choice theory has been shaped by the formative theorems of Arrow and Gibbard & Satterthwaite. Motivated by the Gibbard-Satterthwaite theorem effectively showing that all sensible voting rules are manipulable i.e., susceptible to strategic voting, metrics comparing the degree of manipulability of different voting rules have been developed. However, the concept of manipulation is often deemed undesirable without explicit justification. The objective of this thesis is to evaluate the ethical status and democratic implications of strategic voting, and what the significance of these factors are for applying the manipulability metrics. In particular, this thesis aims to assess the grounds on which concepts related to strategic voting can justifiably be used to argue for adopting one voting rule over another. A literature review was conducted on the metrics developed for measuring manipulability of voting rules and the welfare impacts of manipulation, as well as the ethical and democratic arguments for and against strategic voting. The findings of the literature review were applied to the discussions on electoral reform in the United States, focusing on the ongoing debate between Ranked-Choice Voting (RCV) and plurality voting. The findings of the literature review show that measuring manipulability is dependent on a variety of assumptions used in the different metrics, thus lacking unified consensus on the manipulability of different voting rules. The ethical status and democratic implications of strategic voting are found to be contestable as well. This implies that for a metric of manipulability to carry prescriptive weight in discussions of real-world election reform, the assumed normative wrong of strategic voting should be explicitly substantiated, and moreover, it should be justified how the used metrics tracks this particular concern.