Modeling the diffusion of homicide by type through a mid-sized American city to inform public health prevention efforts

Purpose: Recent research demonstrated that the spatio-temporal movement of homicide through Newark, New Jersey can be modeled using techniques from the field of medical geography. Homicide, however, is a complex phenomenon and different types of homicide may have different patterns of movement. Identifying these patterns is critical to the deployment of prevention efforts specific to a homicide type. Methods: Homicide data were obtained from the Newark Police Department's Homicide Unit's log sheets and investigative files from 1997 through 2007 (N=590). The address location at which each homicide victim was found was geo-coded, and date and type of homicide recorded. We used cluster detection software to model the spatio-temporal movement of statistically significant homicide clusters by type, using census tract and month of occurrence as the spatial and temporal units of analysis. Results/Outcomes: During the study period, there were 127 homicides that were the result of escalating disputes; 113 homicides that arose from interpersonal disputes; 95 family and intimate partner homicides; 175 homicides related to drugs and gangs; and 80 robbery homicides. Findings on the diffusion or non-diffusion of homicide clusters for each of these homicide types will be presented. Conclusions: There exist promising intervention efforts for many homicide types, but interventions targeting one type may not prevent homicides with different etiologies. By tracking how homicide types diffuse through communities, and determining which places have ongoing or emerging homicide problems by homicide type, we can better inform the deployment of prevention and intervention efforts.