Health and the Environment: A Geographical Study of Drugs at Different School Neighbourhoods

This paper studies drug offence at schools and their environment to explain why some but not all schools had drug offences. We used a geographical approach that integrated spatial data of crime, census, and the built environment to identify potential risk factors of drugs at schools. Based on all recorded drug offences at schools (2001-2008) in the region of York in Ontario, Canada, from the police, we geocoded the schools and analysed the associations between drugs at schools and their environment (i.e., neighbourhood characteristics of schools). Neighbourhood characteristics were represented by census socioeconomic variables (including unemployment, lone parent, residential instability, immigration, adults at home, education, and ethnic heterogeneity) and built environment variables (open space and housing type). The analyses were done using geographical information systems for geocoding and extraction of attributes that represent neighbourhood characteristics. Bayesian statistical methods with Markov chain Monte Carlo simulation were used to fit the logistic regression models adopted. One of the major findings was that drug offence at schools and lone parents in the school neighbourhoods is significantly associated. Our findings have the potential to contribute to relevant policy discussions in reducing drug activities at schools. Limitations of the study and further research are discussed.