By Ahmad Ilderim Tokey

Objectives: This study expands the literature by finding the associations of land use (LU) and road-related Built Environment with property and violent crime in Detroit from 2019-to 2021 and documenting if the relationships have changed during COVID-19. Method: This study uses property and violent crime location data from 2019 to 2021. It then builds Geographically Weighted Regression and Spatial Error Models to formalize the BE-crime relationship. Results: Findings indicate that multifamily LU and liquor stores are related to property crime but not much to violent crime. The retail and office LU proportion, bus stop density, and density of roads less than 40 miles per hour are positively linked with crime rates. Reversely, block groups' median income, population density, and tenure length are inversely associated with crime rates. The local association over the year is consistent for property crime but varied substantially for violent crime. In the low-income neighborhoods, single-family got more associated with violent crime during the pandemic while in a high-income neighborhood, single-family LU revealed a negative association with violent crime. Further, areas, where retail LU had a strong positive relationship with violent crime rates in 2019, attracted more crime afterward. The high bus stop densities downtown got more positively associated with violent crime in 2020-21 than in pre-pandemic time. Conclusion: This study advances understanding related to the BE-crime relationship, sheds new light on street-related BE, and documents the links between BE and crime. For local

Columbus, OH: Ohio State University, Department of Geography, 2022. 38p.