Identifying situational predictors of police shootings using multivariate analysis

Prior research on the police decision to use deadly force has tended to neglect multivariate relationships, particularly at the situational level. This paper makes use of data describing deadly force incidents in Philadelphia during two time periods (1970‐1978 and 1987‐1992) and employs multivariate analyses to identify situational predictors of police shootings involving gun‐assaultive suspects. Findings from the multivariate analyses are then used in a pilot effort to develop predictive risk classifications of deadly force incidents. Identification of predictors of deadly force is helpful not only in assessing the relative contributions of situational variables but also in shaping our understanding of the behavior of line officers who are forced, by the nature of their work, to make split‐second decisions involving life and liberty with minimal guidance and support from the police department.