The intersection of nature, segregation, and police violence : a multidimensional study exploring built environment pathways in deprived U.S. neighborhoods

Abstract

This dissertation offers a comprehensive and systematic analysis of the complex interplay between critical built environment determinants and police violence, including its racial disparities, within the United States. Built on solid theoretical frameworks and empirical evidence, the study employs the multiscale approach to examine how green spaces and multidimensional residential segregation influence both lethal and non-lethal police violence. The research also further examines how such associations perform within metropolitan areas and across different levels of social disparities. Chapter I presents a systematic literature review following the PRISMA guidelines to understand the existing theories and empirical evidence, identify the potential built environment factors, and to propose the research gaps and questions. Following that, Chapter II, the preliminary study, explores the relationship between greenness levels and fatal police shootings across 3,108 U.S. counties from 2016 to 2021. Using the BYM Poisson regression model that accounts for spatial autocorrelation, the study reveals a significant negative association between greenness and fatal police shootings. A more critical finding is such association remains stronger in metropolitan counties and has a more pronounced effect in areas with higher social deprivation score (SDI). These findings reaffirm the potential effects of green spaces and the built environment factors, providing a foundation for subsequent analyses in the following chapters. Chapter III explores the differential impacts of specific green space types and their associations with fatal police shootings. Building on findings of Chapter I, the study examines the relationship between detailed green space types with police shooting cases over a five-year period across all 3,108 counties in the United States at the county level. A multi-layer Besag-York-Mollié (BYM) model is employed controlling for socioeconomic, demographic, law enforcement, and other environmental factors. The study demonstrates that open spaces, forests, and agricultural lands are significantly and negatively associated with reductions in fatal police shootings at the county level. Chapter IV shifts focus to residential segregation, a well-established determinant in the existing theories and empirical evidence. The study explored particularly how multidimensional residential segregation influences the Black-White racial disparities in both lethal and non-lethal police violence. The study employes four hierarchical GRM regression models to analyze the tract-level associations. The findings reveal that residential segregation significantly and independently associates with racial disparity of police violence. Different dimensions of segregation have different effects. Among the four segregation dimensions, Exposure shows the most substantial negative effect, with Concentration and Clustering also demonstrating significant associations. Key contributions of this dissertation include the preliminary introduction of an environmental framework for understanding police violence, the identification of specific built environment factors associated with violence in deprived areas. Also, for the first time, this research introduced the Social Deprivation Index (SDI) as a moderating variable in each Chapter, confirming that the impacts of built environments on reducing police violence were more significant among socioeconomically disadvantaged groups. The study suggested that improving the urban built environment—especially by increasing green spaces and optimizing residential structures—is an effective strategy for reducing police violence and its racial disparities.