Shaily Krishan1, Roger Bakeman2, Beth Broussard3, Sarah L Cristofaro1, Dana Hankerson-Dyson1, Letheshia Husbands1, Amy C Watson4, Michael T Compton5. 1. Emory University School of Medicine, Department of Psychiatry and Behavioral Sciences, Atlanta, GA, USA. 2. Georgia State University, Department of Psychology, Atlanta, GA, USA. 3. Lenox Hill Hospital, Department of Psychiatry, New York, NY, USA. 4. University of Illinois at Chicago, Jane Addams College of Social Work, Chicago, IL, USA. 5. Lenox Hill Hospital, Department of Psychiatry, New York, NY, USA; Hofstra North Shore - LIJ School of Medicine at Hofstra University, Department of Psychiatry, Hempstead, NY, USA. Electronic address: mcompton@nshs.edu.
Abstract
OBJECTIVE: Police officers' decisions and behaviors are impacted by the neighborhood context in which police encounters occur. For example, officers may use greater force and be more likely to make arrests in disadvantaged neighborhoods. We examined whether neighborhood characteristics influence police encounters with individuals suspected to have a serious mental illness, addictive disorder, or developmental disability. METHOD: We obtained data on 916 encounters from 166 officers in six jurisdictions in Georgia, USA and abstracted geographical data pertaining to the location of these encounters from United States Decennial Census data. Encounters were nested within 163 census tracts. Officer-reported data covered general encounter characteristics, the officer's perception of the subject's condition, subject demographics, use of force, and disposition of the encounter (e.g., arrest v. referral or transport to treatment services). Geographical data included 17 variables representing population and housing characteristics of the census tracts, from which three indices pertaining to neighborhood income, stability, and immigration status were derived using factor-analytic techniques. We then examined associations of these indices with various encounter-related variables using multi-level analysis. RESULTS: Encounters taking place in higher-income and higher-stability census tracts were more likely to be dispatch-initiated and take place in a private home compared to those in lower-income and lower-stability neighborhoods. In higher-income neighborhoods, encounters were more likely to involve a subject suspected to have a mental illness (as opposed to an addictive disorder or developmental disability) and less likely to involve a subject suspected to have alcohol problems. The officer's level of force used was not associated with neighborhood factors. Regarding disposition, although the likelihood of arrest was unrelated to neighborhood characteristics, encounters taking place in higher-immigrant neighborhoods were more likely to result in referral or transport to services than those in lower-immigrant neighborhoods. CONCLUSION: Neighborhood characteristics are important to consider in research on police interactions with individuals with serious mental illnesses, addictive disorders, or developmental disabilities. Such research could inform departmental training policies and procedures based on the needs of the jurisdictions served.
OBJECTIVE: Police officers' decisions and behaviors are impacted by the neighborhood context in which police encounters occur. For example, officers may use greater force and be more likely to make arrests in disadvantaged neighborhoods. We examined whether neighborhood characteristics influence police encounters with individuals suspected to have a serious mental illness, addictive disorder, or developmental disability. METHOD: We obtained data on 916 encounters from 166 officers in six jurisdictions in Georgia, USA and abstracted geographical data pertaining to the location of these encounters from United States Decennial Census data. Encounters were nested within 163 census tracts. Officer-reported data covered general encounter characteristics, the officer's perception of the subject's condition, subject demographics, use of force, and disposition of the encounter (e.g., arrest v. referral or transport to treatment services). Geographical data included 17 variables representing population and housing characteristics of the census tracts, from which three indices pertaining to neighborhood income, stability, and immigration status were derived using factor-analytic techniques. We then examined associations of these indices with various encounter-related variables using multi-level analysis. RESULTS: Encounters taking place in higher-income and higher-stability census tracts were more likely to be dispatch-initiated and take place in a private home compared to those in lower-income and lower-stability neighborhoods. In higher-income neighborhoods, encounters were more likely to involve a subject suspected to have a mental illness (as opposed to an addictive disorder or developmental disability) and less likely to involve a subject suspected to have alcohol problems. The officer's level of force used was not associated with neighborhood factors. Regarding disposition, although the likelihood of arrest was unrelated to neighborhood characteristics, encounters taking place in higher-immigrant neighborhoods were more likely to result in referral or transport to services than those in lower-immigrant neighborhoods. CONCLUSION: Neighborhood characteristics are important to consider in research on police interactions with individuals with serious mental illnesses, addictive disorders, or developmental disabilities. Such research could inform departmental training policies and procedures based on the needs of the jurisdictions served.
Authors: Michael T Compton; Roger Bakeman; Beth Broussard; Dana Hankerson-Dyson; Letheshia Husbands; Shaily Krishan; Tarianna Stewart-Hutto; Barbara M D'Orio; Janet R Oliva; Nancy J Thompson; Amy C Watson Journal: Psychiatr Serv Date: 2014-04-01 Impact factor: 3.084
Authors: Michael T Compton; Roger Bakeman; Beth Broussard; Dana Hankerson-Dyson; Letheshia Husbands; Shaily Krishan; Tarianna Stewart-Hutto; Barbara M D'Orio; Janet R Oliva; Nancy J Thompson; Amy C Watson Journal: Psychiatr Serv Date: 2014-04-01 Impact factor: 3.084