Measuring disparities in police use of force and injury among persons with serious mental illness

OBJECTIVES: To measure disparities in experience of police use of force and injury among persons with serious mental illnesses. METHODS: We gathered novel police use of force and suspect injury data from 2011 to 2017 from a nonrandom sample of nine police departments in the United States and used synthetic methods to estimate the share of the local population with serious mental illness. We estimate disparities using multi-level models estimated in a Bayesian framework. RESULTS: Persons with serious mental illness constitute 17.0% of use of force cases (SD = 5.8) and 20.2% of suspects injured in police interaction (SD = 9.0) in sample cities. The risk that persons with serious mental illness will experience police use of force is 11.6 times higher (95% CI, 10.7–12.6) than persons without serious mental illness. Persons with serious mental illness are also at a higher risk of experiencing injury, 10.7 times (95% CI, 9.6–11.8), relative to persons without serious mental illness. These relative risk ratios are several times larger than racial and ethnic disparities estimated in the same cities. CONCLUSION: Persons with serious mental are at a significantly elevated risk of experiencing police use of force and injury in police encounters than the general public. The disparities we estimate are several times higher than racial/ethnic disparities in force and injury. Efforts to reform police practices and reimagine public safety in the United States should address significant disparities in police use of force against those with serious mental illness. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12888-021-03510-w.