Some environmentally relevant goods may also be positional (e.g., car owners may get utility from having a better car than others), which will influence optimal taxation (e.g., Frank 1985). If there is habit formation, myopia, or even addiction, then the complexity of the utility function increases more (Rabin 1998; Becker and Murphy 1988), which may or may not have consequences for policy instrument design. If policymakers maximize a function completely different from welfare (e.g., chance of being reelected) or if they follow some other rationale for decisionmaking (e.g., psychological factors, ideology, or norms), then it must be incorporated into the analysis, as is traditionally done in the public choice literature-particularly if the purpose of the analysis is to be positive rather than normative. In this area, economists tend to recommend “optimal” policies (mainly based on efficiency criteria) and thus slip into a rather normative role. Equally as important as designing policies should be analyzing and understanding why certain policies are selected or designed differently under different circumstances. Psychological, cultural, and political factors may dominate over economic ones.