The potential for violence in arrests of persons with mental illness

Purpose: The purpose of this paper is to determine whether police use of force and suspect resistance are more likely to occur in arrest encounters involving suspects with mental health problems. Design/methodology/approach: The study uses data from interviews with 942 individuals recently arrested by officers in more than a dozen different police departments in Maricopa County, Arizona in 2010. Both logistic and ordinal regression analyses are used to predict two models of suspect resistance (resistance in the current arrest, resistance in a previous police contact) and three models of police use of force (any force in the current arrest, ordinal measure of force in the current arrest, and any force in a previous contact). Findings: The results provide empirical support for a link between mental illness and increased resistance against the police. With regard to arrestee mental illness and use of force, the results are mostly consistent with prior research suggesting a null relationship, with an important caveat involving greater use of higher level, weapon force. Research limitations/implications: The study suffers from the traditional limitations associated with self-report data, and the generalizability of the findings beyond arrest encounters in Maricopa County is not known. The explanatory power of the multivariate models was relatively weak, suggesting a good degree of unexplained variance. Practical implications: The non-significant relationship between arrestee mental illness and use of force is consistent with efforts by police to improve their response in these complex encounters. The significant weapon-force finding may suggest that police respond to the affronts of mentally ill suspects differently than affronts from other suspects. The non-significance of key extra-legal factors suggests that police decisions to use force were not influenced by arrestee race/ethnicity, age, or social standing. Originality/value: Unlike previous studies, the current research uses self-reported measures of mental health problems. The current study also examines arrests from more than a dozen different police departments.