Representation, reputation and attitudes of policy audiences

Abstract

This thesis examines how reputational beliefs can become a boundary condition for symbolic representation effects. Exploring the literature about representative bureaucracy, symbolic representation and bureaucratic reputation, we ask how audiences that share positive or negative reputational beliefs about a bureaucracy evaluate its future performance when they perceive the bureaucracy is highly representative at leadership levels. After that, we ask how distinct audiences (one close to the bureaucracies, others distant from it) would react given the same issues. We relied on two survey experiments and qualitative data to advance theory about the interaction of bureaucratic reputation and symbolic representation. One survey experiment tested our hypotheses in two bureaucratic agencies: police, about whom audiences share negative reputational beliefs, and public schools, characterized by positive reputational beliefs among policy audiences. The second survey experiment focused on two distinct audiences, slum dwellers (target audiences of the given agencies) and outsiders from a middle-class neighborhood (distant evaluators to both agencies). Our results demonstrate the direct effects of representation and reputation for two bureaucratic performance variables: procedural justice and organizational effectiveness. In our first experiment, we also observed an interaction effect between representation and reputation for one of the bureaucratic performance variables, procedural justice – expectations audiences share about their future treatment. When comparing slum dwellers with outsiders, we also observed that the distant evaluators (outsiders) had in general a lower perception of reputation and could not perceive the same interaction effect felt for slum dwellers. Additional qualitative analysis of justifications given by participants was used to make sense of the effects encountered in this endeavor. Results demonstrated that slum dwellers were more likely to hold positive expectations for public schools regardless of representation levels and that outsiders were more likely to behave in an uninformed way, overestimating representational effects and underestimating reputational ones. Qualitative analysis also revealed that the effects of representations are stronger for procedural justice, because the leader has more discretion in the form the agency treats local citizens. However, respondents recognize that the overall improvement of the agency lies beyond leaders` discretion.