OBJECTIVES: To explore the cross-sectional association between alcohol outlet density and police events in Manukau City, New Zealand. METHODS: Using data for the Census Area Unit (suburb) level, per-capita measures of alcohol outlet density for January 2009 were calculated for off-licence outlets, clubs and bars, and restaurants and cafés. Data on police events and motor vehicle accidents were obtained for the period 1 July 2008 to 30 June 2009, and also converted into per capita measures. A spatial seemingly unrelated regression model was developed, which simultaneously assessed the relationship between densities and all nine categories of police events, and motor vehicle accidents, while controlling for relevant covariates. RESULTS: All three outlet density measures were significantly associated with a range of police events, but only off-licence density was significantly associated with motor vehicle accidents. An additional off-licence outlet in a given area was associated with 85.4 additional police events and 10.3 additional motor vehicle accidents; an additional club or bar was associated with 34.7 additional police events and 0.5 additional motor vehicle accidents; and an additional restaurant or cafe was associated with 13.2 additional police events and 2.1 additional motor vehicle accidents. CONCLUSIONS: The results do not imply causality. However, they are broadly consistent with availability theory, and imply that local alcohol policy should account for the effects of additional outlets when new licences are granted. While the methodological approach described here is easily transferable to investigate the relationships elsewhere, we suggest some areas for improvement of future studies.
OBJECTIVES: To explore the cross-sectional association between alcohol outlet density and police events in Manukau City, New Zealand. METHODS: Using data for the Census Area Unit (suburb) level, per-capita measures of alcohol outlet density for January 2009 were calculated for off-licence outlets, clubs and bars, and restaurants and cafés. Data on police events and motor vehicle accidents were obtained for the period 1 July 2008 to 30 June 2009, and also converted into per capita measures. A spatial seemingly unrelated regression model was developed, which simultaneously assessed the relationship between densities and all nine categories of police events, and motor vehicle accidents, while controlling for relevant covariates. RESULTS: All three outlet density measures were significantly associated with a range of police events, but only off-licence density was significantly associated with motor vehicle accidents. An additional off-licence outlet in a given area was associated with 85.4 additional police events and 10.3 additional motor vehicle accidents; an additional club or bar was associated with 34.7 additional police events and 0.5 additional motor vehicle accidents; and an additional restaurant or cafe was associated with 13.2 additional police events and 2.1 additional motor vehicle accidents. CONCLUSIONS: The results do not imply causality. However, they are broadly consistent with availability theory, and imply that local alcohol policy should account for the effects of additional outlets when new licences are granted. While the methodological approach described here is easily transferable to investigate the relationships elsewhere, we suggest some areas for improvement of future studies.