Behavioral science to strengthen democracy

Abstract

Democracy faces critical challenges, including declining trust, misinformation, political polarization, and the strain on election systems. Despite its importance, behavioral science has contributed relatively little to addressing these issues compared to fields like health and sustainability. This thesis seeks to fill that gap by exploring how behavioral science can strengthen democracy through practical, evidence-based interventions. Chapter one reviews the evidence and opportunities. Chapter two demonstrates how operational transparency, using video, print, and email interventions, increases voter trust and turnout, especially among those who have no party preference. Chapter three examines the role of trusted messengers and inoculation techniques, such as prebunking and accuracy nudges, in combating misinformation. Findings reveal that local, trusted figures, like election officials and pastors, effectively improve voter trust and truth discernment, particularly during active election periods. Chapter four investigates decision aids in participatory budgeting, showing that these tools promote more equitable voting outcomes but reveal a disconnect between voter perceptions and behavior. The findings underscore the importance of refining ballot design to improve decision-making quality. Together, these papers highlight the untapped potential of behavioral science to address democratic challenges. In the general discussion, chapter five, this thesis calls for incremental experimentation, interdisciplinary collaboration, and scaling evidence-based interventions to ensure democracy remains resilient. By contributing to a growing body of research, this thesis advocates for “behavioral democracy,” a systematic effort to integrate behavioral science into election governance. Strengthening democracy is urgent and essential for collective human flourishing, and this work offers a foundation for meaningful progress.