|Recent research of police workload in Anchorage, Alaska has shown that patrol officers were able to detect the general contours of alcohol-crime and drug-crime assocations for incidents to which they responded (Myrstol, Schafer and Giblin 2003). Patrol officer expectations that violent incidents were more likely to be alcohol-related than drug-related were supported. Supplementary incidents logs completed by patrol officers also lent support to patrol officers' perceptions that property offenses were more likely to e linked with illicit drugs than with alcohol. However, the magnitude of alcohol's association with violent incidencts, and drugs' relationship with property offense calls, was shown to be significantly less than what officers expected. That is, the group of patrol officers studied by Myrstol et al. was found to have fairly exaggerated views of the link between alcohol, rugs, and crime.
The perceptual "gap" shown to exist between patrol officer perceptions of alcohol and drug involvement in problems of crime and social disorder and the empirical (observed) reality of patrol work raises several theoretical, operational, and policy questions, one of which is "How do officers decide that alcohol or drugs are involved in an incident?" This paper builds on the exploratory study conducted by Myrstol, Schafer and Giblin. For each alcohol- and/or drug-related incident recorded during the study period, officers recorded how they came to that determination: a) direct observation; b) visible impairment; c) detectable odor; d) third-party information; e) prior knowledge of involved party; f) paraphernalia present; g) direct inquiry; h) breath/sobriety test administration; i) other. An attempt is made to statistically model the dimensions of these patrol officers' determinations of alcohol and/or drug involvement, and the impacts of such determinations on the decision to arrest.
(Return to Program Resources)