Noncompensatory Rules, Voting, and Welfare

Abstract

Individual decision behavior has been described by several authors as noncompensatory: decisions are based on just a subset (often a minimal subset, i.e. one) of the dimensions that could be thought relevant. A further claim that has been advanced is that these rules are adaptive (that they “make us smart”), because the environments in which decisions are made are themselves noncompensatory -- affording the discard of most information. This paper examines the consequences of one commonly studied type of noncompensatory rule -- lexicographic choice -- applied in the context of voting, both by individuals and collectives. Theoretical and empirical arguments are put forward which question the collective efficacy of noncompensatory rules, and lexicographic voting specifically, while suggesting that they may yet be descriptive and explanatory of voting behavior and political party positioning as it evolves across multiple elections. Lexicographic voting provides one model for how political parties (e.g. the Republicans and Democrats in the United States) could essentially swap electoral bases or platforms over the course of many elections.