Modeling Public Decision Preferences Using Context-Specific Value Hierarchies

Abstract

Although a universal hierarchy of public values has proven elusive, the literature in individual-level values suggests that decision makers do organize their personal values into hierarchies based on context. Through analysis of public values and public decision preferences gathered in a pilot study of city-level public administrators ( n = 182), we use an empirical approach to identify context-relevant public values for five different decision contexts. We then demonstrate multiple possible approaches to modeling individual- and community-level policy preferences based on value hierarchies derived from the individual-level data. We find that the predictions based on value hierarchies are better than would be predicted in the absence of such hierarchies, and that these differences are statistically significant. These findings suggest that while creating a universal hierarchy of values remains challenging, context-relevant public value hierarchies at smaller units of analysis may be useful in describing, predicting, and explaining the decisions of public administrators.