DRINKING IN VIOLENT AND NONVIOLENT EVENTS LEADING TO ARREST - EVIDENCE FROM A SURVEY OF ARRESTEES

The minimal requirements for evaluating the relative magnitude of acut e and chronic effects of alcohol on violent crimes include: (a) a sele ction of comparable samples of violent and nonviolent events, and (b) adequate controls for the anticipated positive correlation between dri nkers and drinking in feasible samples. Following these requirements, hypotheses about the impact of personal drinking patterns and drinking in events on violent crimes in a random sample of 1,147 arrestees int erviewed in a California detection facility in 1989 were tested. Using multiple indicators of drinking in events, and taking into account er rors of misclassification and correlations between drinkers and drinki ng, it was found that arrestees charged with violent crimes showed sig nificantly higher rates of drinking in events related to arrest. On th e other hand, the twelve-month drinking pattern of arrestees did not d iscriminate between violent and nonviolent charges in this sample, giv en controls for drinking in events.