Research on mental illness in relation to social problems such as crime, unemployment, and homelessness often ignores the broader social context in which mental illness is embedded. Policy, research, and practice will be improved if greater attention is given to social context. The authors critically analyze the approach used in much of the psychiatric services literature to infer links between mental illness and social problems. They compare these studies with studies that have been more validly conceptualized to account for social context. With this broader perspective, the impact of mental illness on crime, unemployment, and homelessness appears to be much smaller than that implied by much of the psychiatric services literature. Poverty moderates the relationship between serious mental illness and social problems. Factors related to poverty include lack of education, problems with employment, substance abuse, and a low likelihood of prosocial attachments. This relationship is often complicated and is not amenable to simple explanations. Research and policy that take this complexity into account may lead to greater effectiveness in interventions for persons with serious mental illness.

The Working Alliance Inventory was used to measure the strength of the therapeutic relationship between seriously mentally disabled case management clients and their case managers in a randomized trial of consumer-provided case management services. It was found that while there was no difference in the strength of the alliance between the consumer and nonconsumer teams of case managers, there were positive relationships between alliance and some outcomes, including quality of life, symptomatology, attitudes toward medication compliance, and satisfaction with mental health treatment.

The Crisis intervention team model (CIT) is possibly the most well known and widely adopted model to improve police response to persons with mental illness. A primary goal of CIT programs is to divert individuals with mental illness from the criminal justice system to mental health services. In this paper we examine the effectiveness of fielding CIT trained and supported officers for influencing call outcomes using data from patrol officers (n = 112) in four Chicago Police districts. Results from regression analysis indicate that CIT certified officers directed a greater proportion of persons with mental illness to mental health services than their Non-CIT certified peers. CIT did not have an immediate effect on arrest. Moderator analysis indicates that CIT had its biggest effect on increasing direction to services and decreasing "contact only" among officers who have a positive view of mental health services and who know a person with mental illness in their personal life. Additional moderators of the CIT effect on call outcomes include level of resistance and the presence of a weapon. Findings from this study have important implications for policy, practice and future research.

The large numbers of people with mental illness in jails and prisons has fueled policy concern in all domains of the justice system. This includes police practice, where initial decisions to involve persons in the justice system or divert them to mental health services are made. One approach to focus police response in these situations is the implementation of Crisis Intervention Teams (CIT). The CIT model is being implemented widely, with over 400 programs currently operating. While the limited evidence on CIT effectiveness is promising, research on CIT is limited in scope and conceptualization-much of it focusing on officer characteristics and training. In this paper we review the literature on CIT and present a conceptual model of police response to persons with mental illness that accounts for officer, organizational, mental health system and community level factors likely to influence implementation and effectiveness of CIT and other approaches. By moving our conceptualizations and research in this area to new levels of specificity, we may contribute more to effectiveness research on these interventions.