Identifying Dangerous Places to Homeless People: An Application of Risk Terrain Modeling

Spatial analysis in criminology is an emerging area due to the rapid development of technology. The spatial analysis can forecast future crime locations by considering prior crime and risk factors that potentially lead to crime. However, most study in environmental criminology has exclusively conducted on the relationship between place and crime. Thus, this thesis starts with selecting a specific individual group, which is vulnerable to crime, to make a better analysis. Also, given that there is no prior study on the risk terrain modeling (RTM) to the homeless related crime, exploring the spatial influence of the identified risk factors within the city of Los Angeles might help to inform a potential solution to the criminal justice system. The results show that homeless people are vulnerable to crime at 22 out of 26 identified specific locations regardless of types of crime. Specifically, homeless individuals tend to commit a crime or be victimized by crime near past drug arrest locations and homeless shelters. These findings explain why homeless-related crimes occur in specific areas with more micro-level of analysis. Also, the results prove that selecting a particular group at the beginning stage in the spatial crime analysis can be a promising approach to reducing the crime.