This study includes several separate analyses that used data from a 1995 study of 1,585 arrests made in Phoenix in 1994 and both replicated and reversed the methods of the original study to determine the likely predictors of police use of force in making an arrest, as well as use of force by the suspect.
Results of the reanalysis revealed that consistent predictors of police use of force included the suspect’s use of force, the number of police officers initially at the scene, a change in the number of police, and police use of the contact and cover tactic. Other predictors included the presence of bystanders at the arrest, lack of visibility at the arrest scene, a suspect known to be assaultive or resistive or carrying a weapon, and a male suspect. Consistent predictors of a suspect’s use of force included a violent offense that did not involve domestic violence, a vice offense, a domestic violence offense, the presence of bystanders, impairment of the suspect by alcohol, the suspect’s association with a gang, the suspect’s younger age, and the suspect being non-Hispanic. Findings indicated that limiting the number of police officers at the scene was advantageous in some situations and suggested that police officers should use calm, nonthreatening tones of voice to increase the likelihood of suspect compliance. The analysis concluded that police were more likely to cause suspects to use force as suspects were to cause police to use force. Figures, tables, appended instrument and tables, and 23 references
Date Published: January 1, 1999