UCLA

Policymakers are increasingly using interventions, informed by psychological science, to motivate behavior change (“behavioral policies”) across a range of domains: from reducing failure to appear in court to increasing COVID19 vaccinations. Field experiments testing these behavioral policies are theoretically and practically important: they provide ecological validity to psychological theories, while evaluating real solutions to pressing social problems. Yet, recent evidence suggests, when tested in interventions in the field, promising ideas from psychology often yield null effects or even backfire. Moreover, behavioral policies testing the same psychological principle often result in divergent results across field experiments. This presents multiple obstacles for behavioral science. Without theories that can explain unexpected and divergent evidence across field experiments, researchers may prematurely conclude that psychological principles that have been theorized to be highly motivating do not have ecological validity, and policymakers may prematurely conclude that behavioral policies are not useful tools for addressing problems at scale. While recent evidence has pointed to differences in sampling and measurement to explain unexpected and divergent results, my dissertation explores a different hypothesis: people draw inferences about policymakers from details in the design of behavioral policies and the contexts in which they are implemented, which can help determine when policies work as intended and when they backfire. My dissertation includes four chapters, each highlighting an inference people draw from behavioral policies: inferences about urgency, support, inauthenticity, and trustworthiness. By identifying social inferences as a hidden mechanisms underlying the effects of behavioral policies, my research provides new insight into (1) the unexpected effects of behavioral policies that both academics and expert practitioners previously believed to be effective, (2) the redesign of psychological interventions so that they have their intended positive effects, (3) the identification of moderators that help explain the divergent effects of interventions across contexts, and (4) the inclusion of consequential yet unexplored outcomes in the evaluation of behavioral policies. Together, my dissertation integrates and advances knowledge on social inference making and behavior change, provides a new perspective on studying psychological theories in the field, and offers practical implications for policymakers seeking to design effective behavioral policy.