Ethnographic Methods in Research on Homeless Services: Challenges of Integration and Interpretation (Part 2-Case Managers)

Introduction: This two-part report discusses the use of diverse sources of observational and interview data in a longitudinal qualitative study of mental health recovery among formerly homeless individuals with serious mental illness enrolled in supported housing programs in New York City. In addition to in-depth interviews with clients and their case managers, we used: 1) participant observation of program operations (site visits, ride-alongs on home visits) and of participants’ daily routines (walk-along interviews); and 2) respondent-controlled photo elicitation interviewing (PEI).  This second of two papers focuses on the program site and case manager data and preliminary findings. Research questions guiding this inquiry include: 1) how are multiple techniques of qualitative data collection coordinated using ethnographic methods? 2) how complementary are these data sources during analysis and interpretation? 3) how are multi-method qualitative findings interpreted to optimize theory development?

Methods:  Two supported housing programs in New York City were studied and 34 case managers (CMs) were interviewed with their client’s consent.  Participant observation was conducted of staff meetings, community (client) meetings, and ride-alongs for apartment visits. Field notes, photographs and mapping or diagrams provided visual data for analysis. Open and thematic coding of the in-depth interviews was supplemented by the observational data to examine barriers to and facilitators of recovery practices in the program’s operations. Specific attention was given to contrasts between interviews and observation.

Findings:  Interview data did not contradict observation but the latter revealed tensions between staff and clients focused on the restrictions placed on clients (apartment rules, long waits for independent housing, limited contact with staff, sobriety expectations) and staff role expectations. Indeed, preliminary observation led to more focused interview questions about practice routines and supervision. Overall themes from the analyses include: security concerns, surveillance and privacy, temporariness and uncertain futures. Concerns for security (of the offices, in the apartments, and during case manager-client interactions) were pervasive, the case manager’s role in surveillance was observed as was the client’s lack of privacy. Time spent on housing applications and staff turnover undermined engagement and the temporary nature of housing and services eroded optimism about the future. Analyses also revealed that clinical or therapeutic relationships were limited by managerial duties (paperwork, supervision of clients) and staff turnover.

Discussion: The nature of the client-program relationship was strained by the demands of supervision, rules monitoring and managerial work. Organizational and ‘street level bureaucrat’ theories only partially explain this phenomenon. The findings link ‘up and out’ to larger issues of resource scarcity in homeless services and a lack of affordable housing resulting in the dependence of clients on restrictive programs for indefinite periods of time. The complementary rather than contradictory aspects of observational and interview data led to a far more detailed grasp of staff practices and program operations as they interfaced with clients’ needs and expectations.