Competing Concerns in Efforts to Reduce Criminal Legal Contact Among People with Serious Mental Illnesses: Findings from a Multi-City Study on Misdemeanor Arrests.

People with serious mental illnesses are disproportionately involved in the criminal legal system, often for low-level, non-violent misdemeanors. This paper examines how decision-makers at different stages of the criminal legal system articulate unique visions of the "best approach" for addressing this problem of over-representation. Focus groups and in-depth interviews were conducted with 94 stakeholders from Atlanta, Chicago, New York City, and Philadelphia to understand how decision-makers from different agencies use and process specific misdemeanor charges in relation to people with serious mental illnesses. Data were analyzed using a thematic approach. The data reveal a series of tensions regarding how criminal legal system stakeholders process people with serious mental illnesses through the misdemeanor system. Three key themes emerged from analysis. The first characterizes the shared commitment across agencies to reducing system contact among people with mental illnesses. The second explores how agencies differ on how to make good on that commitment because of the distinct values and goals they bring to the table. The final theme explores the limits of current approaches to reducing system contact for people with mental illnesses. Findings are discussed in the context of literature on "loose coupling" and the focal concerns framework and demonstrate that decisions about how and when to intervene with people with mental illnesses in the criminal legal system are influenced by the varying orientations, goals, and values of stakeholder agencies. Understanding these core differences is a critical step toward value alignment in strategies to reduce system involvement among people with mental illnesses.

Misdemeanor Charges Among Individuals With Serious Mental Illnesses: A Statewide Analysis of More Than Two Million Arrests.

OBJECTIVE: Reducing the overrepresentation of individuals with serious mental illnesses in the criminal legal system requires a better understanding of the charges for which they are most commonly arrested. This study aimed to compare violent offenses, penal code classifications, Uniform Crime Reporting (UCR) codes, and specific charges in arrests among individuals with and individuals without serious mental illnesses. METHODS: The authors analyzed all arrests (N=2,224,847) in New York State during 2010-2013. Medicaid data and the state mental health authority's records were used to create an indicator of serious mental illness for each arrest. RESULTS: Among arrests involving individuals with the serious mental illness indicator (N=91,363), 7.3% were for violent offenses, compared with 7.6% of arrests involving individuals without the indicator. Among 10 penal code classifications, class B felonies and class A misdemeanors were more likely in arrests among those with the indicator than among those without it. Of the 14 UCR codes examined, seven were more common in arrests with the serious mental illness indicator. Criminal trespass was among the most common charges in arrests involving individuals with the indicator. CONCLUSIONS: Most arrests involving people with serious mental illnesses were for misdemeanors, specifically class A misdemeanors, and this class comprised a larger proportion of arrests for those with the indicator than of arrests for those without it. New approaches are needed to address the situations-usually related to socioeconomic disadvantage-that result in individuals with mental illnesses receiving misdemeanor charges and cycling through the criminal legal system.

Contexts shaping misdemeanor system interventions among people with mental illnesses: qualitative findings from a multi-site system mapping exercise.

BACKGROUND: People with mental illnesses are disproportionately entangled in the criminal legal system. Historically, this involvement has resulted from minor offending, often accompanied by misdemeanor charges. In recent years, policymakers have worked to reduce the footprint of the criminal legal system. This paper seeks to better understand how misdemeanor systems intervene in the lives of people with mental illnesses. METHODS: System mapping exercises were conducted with misdemeanor system stakeholders from the jurisdictions of Atlanta, Chicago, Manhattan, and Philadelphia. Narrative detail on decision-making and case processing, both generally and in relation to specific types of behavior, including trespassing, retail theft/shoplifting, and simple assault, were coded and analyzed for thematic patterns. Based on the qualitative analysis, this paper offers a conceptual diagram of contexts shaping misdemeanor system interventions among people with mental illnesses. RESULTS: All four sites have been engaged in efforts to reduce the use of misdemeanor charges both generally and in relation to people with mental illnesses. Decision-makers across all sites experience contexts that shape how, when, and where they intervene, which are: (1) law and policy environments; (2) location of the behavior; (3) expectations of stakeholders; (4) knowledge of mental illnesses; and (5) access to community resources. Law and policy environments expand or constrain opportunities for diversion. The location of offending is relevant to who has a stake in the behavior, and what demands they have. Clinical, experiential, and system-level knowledge of mental illnesses inform a chain of decisions about what to do. The capacity to address mental health needs is contingent on access to social services, including housing. CONCLUSION: People making decisions along the criminal legal continuum are critical to illuminating the dynamic, inter-related contexts that facilitate and frustrate attempts to address defendants' mental health needs while balancing considerations of public safety. Multi-sector, scenario-based or case study exercises could help identify concrete ways of improving each of the contexts that surround whole-of-system decisions.