This article develops and tests a comprehensive framework explaining the decision to measure performance, specifically the decision of local governments to conduct citizen surveys. It is structured around a fundamental distinction between subjective performance measures obtained for use in decision making and those that are produced solely for their symbolic value. The author suggests that the field of public administration may be taking a simplified view concerning the promotion and adoption of citizen surveys, overlooking important aspects of the decision-making process of performance-oriented public managers and neglecting the impact of politics and symbols. Political rationality may undercut managerial rationality in the decision to adopt citizen surveys, and symbolic adoption may be the underlying cause of low levels of information use. This study identifies policies to increase adoption of citizen surveys but cautions that simply promoting the adoption of surveys as inherently good may be a naive and wasteful course of action. Practitioners who have already made the decision to measure subjective performance through citizen surveys, or are facing such a decision, can find in this analysis a structure to assess past decisions or guide future decisions.