A Multidimensional Framework for Transportation Safety: Linking Perceptions, Policies, and Environmental Contexts

Abstract

Road safety remains a complex and evolving challenge shaped by behavioral, institutional, and environmental systems. This dissertation integrates multiple analytical perspectives to advance a holistic understanding of how enforcement technologies, workforce readiness, and environmental context collectively influence transportation safety. Through a combination of crash analysis, behavioral research, organizational assessment, and spatial modeling, this dissertation connects technical effectiveness with human and ecological dimensions of safety. The research begins with an evaluation of red-light camera (RLC) enforcement in Phoenix, combining crash data analysis with professional perspectives to assess the program’s operational and safety impacts. Using a before-during-after design with Empirical Bayes estimation, the findings reveal substantial reductions in angle crashes and severe injury during operation. These safety gains remained largely stable even after the program’s termination indicating lasting behavioral adaptation and long-term benefits of enforcement visibility. Building on this evidence, the next phase transitions from professional assessments and data analysis to public perceptions, examining how fairness, transparency, and safety beliefs shape acceptance of RLCs in Arizona and New York. Structural equation modeling demonstrates that fairness and clarity mediate the relationship between safety beliefs and support, underscoring the central role of legitimacy and communication in sustaining automated enforcement programs. The analysis then shifts toward technological transitions and organizational readiness through an examination of electric vehicle (EV) adoption among Minnesota’s public agencies. Surveys and interviews highlight that while environmental commitment and policy alignment foster optimism, persistent concerns about charging reliability, maintenance, and cold-weather performance continue to constrain large-scale implementation. Extending this focus on sustainability and safety, the final investigation emphasizes that transportation safety is not limited to human protection alone. Applying a grid-based spatial model of wildlife-vehicle collisions in Tucson, Arizona, the investigation reveals that roadway density, travel speed, and population exposure are the strongest predictors of crash likelihood, demonstrating how human infrastructure shapes risks for both people and wildlife. Together, these studies form an integrated framework linking behavioral legitimacy, institutional capacity, and ecological context. The findings contribute empirical and methodological insights that advance the design of transportation systems that are safer, more adaptive, and sustainable for all users.