Systematic deviations from rational decision-making in public administration

When making judgments and decisions, individuals have been widely shown to be prone to cognitive biases, leading them to inconsistently reveal their preferences. Despite the fact that these can have far-reaching consequences for our society, research on how systematic errors impinge on the decisions made within the public sector is limited. While a robust discussion has begun on decision-making biases in public management, administration, and policy, these scholars have mostly focused on how citizens make informed assessments of government policies. Only few of them have investigated how public managers and policy makers make decisions. Nonetheless, while exploring the systematic errors of citizens’ judgments about government services is important, it is equally relevant and timely to understand whether and how cognitive biases affect public workers responsible for designing, managing, and implementing public policies and managerial procedures. The three chapters included in this proposal provide empirical evidence, by means of randomized experiments, aimed at increasing our knowledge of these dynamics. In particular, the first chapter replicates and extends previous trials by testing a broad range of cognitive biases on public policy and management decisions on a sample of public managers and employees. While this chapter is purposefully wide-ranging, the remaining two dig deeper on frequent decision-making processes of public workers. More specifically, the second chapter focuses on performance information use by Italian local public managers, investigating how the framing of performance information might differently affect decisions depending on the type of decision that has to be made. The third chapter investigates the micro-foundations of isomorphic pressures which lead to suboptimal decision-making in the public sphere, adding qualitative evidence which helps illuminate the mechanisms behind the causal relations. Public managers and employees who participated in the randomized trials included in chapter 1 were found to be prone to a number of cognitive biases like framing, anchoring, proportion dominance, status quo, and asymmetric dominance. In particular, the framing of outcomes influenced decisions across policy and management domains. In addition, public employees were prone to an anchoring bias when setting standards for responsiveness, were more likely to stick to a suboptimal status quo as the number of superior alternatives increased, and tended to put more effort into activities that affected higher percentages of beneficiaries, even if the absolute number of affected clients was constant. Lastly, decisions changed when a decoy was present, proving an asymmetric dominance effect. In chapter 2, local public managers were more likely to be subject to framing effects under ex post uses of performance information (e.g., service evaluation) than ex ante (e.g., resource allocation). Interestingly, asking them to justify for their choices did not work as a debiasing strategy in their ex post decisions. Finally, the eight experiments included in chapter 3 show that indeed isomorphic pressures can lead to inferior solutions. The qualitative findings help identify public managers and employees who were victims of bandwagon effect and those who gave in to isomorphic pressures because they thought it was the best they could do with the available information.