An Investigation of Drug Offences in Greater Auckland 2005-2011 using Population and Agent-Based Models

Abstract

This thesis uses police offence data to investigate the application of population-based and agent-based modeling to drug offences in the Greater Auckland Area. By investigating common characteristics between the drug offence data and the chosen models it can be shown that the models offer explanatory mechanisms. The investigation takes a stepwise approach moving from the large dataset of 2005 to 2011 through to the individual offender to reflect the process of developing models. Utilizing GIS the large dataset (n=36786) is filtered to show specific drug types; it is on this dataset that the investigations were performed. The first investigation is conducted with a population-based predator-prey model. The model chosen is a model that looks at the adaptive behaviors of prey animals in response to predation. This is linked with the criminological theories of deterrence and diffusion of benefit. The investigation is focused on cannabis offences in the Counties/Manukau district and included the Auckland District of Auckland East. The second investigation uses an agent-based model drawn from mobile communication networks. Mobile Ad hoc Networks (MANETs) model the movement of data across a widely dispersed network of individual devices this characterisation of movement is used to theorise the movement of drugs within drug distribution networks. This investigation focused on methamphetamine supply. It is found in the first investigation that policing did have some local effect on the number of offenders. However use crimes showed a continued upward trend. It is also found that there had been a shift in supply offence from the northern areas to the southern areas of Counties/Manukau district. The second investigation found that the drug supply network did have some common characteristics with the MANET. It is concluded that the policing of cannabis crimes in Counties/Manukau has mixed results when interpreted using population-based approaches and suggests that MANET may prove to be useful in understanding how drugs diffuse amongst social networks and geographic location while remaining tolerant of police attack.