Alcohol Consumption and Homicides in Canada,1950–1999

Abstract

This article addresses whether an association between alcohol consumption and homicide can be established in analyses of Canadian time series data and, if so, whether the strength of the association varies across Canadian provinces and with respect to male and female victim rates. Time series analyses on differenced series of annual aggregate-level data on alcohol sales and homicide rates for the period 1950–1999 were performed for Canadian provinces and the country as a whole. Total alcohol sales were positively and statistically significantly associated with total homicide rates in two provinces and with male homicide rates in three provinces. The effect of alcohol sales was somewhat stronger for male homicide rates than for female homicide rates in two provinces. Pooling of estimates yielded a statistically significant association between alcohol sales and homicide rates for Canada. The findings support the hypothesis that alcohol sales tend to have an impact on homicide rates, and more so in certain provinces and for male homicide rates.