Association for Public Policy Analysis & Management

In recent years, the federal government has shifted its emphasis toward homeless prevention services and stressed the need for an inclusive approach that embraces community partners. This shift parallels the rise in public-private partnerships in which third party organizations, often nonprofits, deliver social services on the government’s behalf. These contractual relationships afford considerable discretion to nonprofit agencies and their frontline workers, but little is known about how nonprofit implementation shapes the services and client experience within a single, well-defined program model. Our research question asks: how does program implementation differ across diverse nonprofit implementing partners and program sites? In this paper, we explore how a local public agency’s funding of a homelessness prevention initiative via contracts with nonprofit organizations affects program implementation.

Our study is situated in the context of the Youth and Family Homelessness Prevention Initiative (YFHPI) in King County, Washington. The program aims to prevent families and unaccompanied youth from becoming homeless through case management and financial assistance. King County contracts with 25 nonprofit partners to administer the YFHPI, which includes dedicated case managers at each nonprofit. The contracted nonprofits vary substantially in their mission and target client populations, ranging from traditional homeless service providers to agencies focused on particular immigrant and refugee populations. Although YFHPI case managers are embedded in each nonprofit provider, the program uses a cohort model where case managers meet regularly for County-organized trainings, professional development, and peer support. The case managers are trained to use a screening tool to verify eligibility for the program based on the household’s vulnerability to homelessness.

We use a qualitative research design to explore how case manager discretion and organizational context affects the services received by clients. Two primary forms of data collection comprise the study. First, approximately 20 hours of direct observation provide a first-hand account of the case managers’ daily activities, including client intake interviews and other case manager work activities, such as professional trainings. Second, 12 to 15 semi-structured individual interviews with case managers offers insight into the practices, rationales, and constraints of their approach to service delivery. Participant selection is conducted using a combination of snowball and purposive sampling methods to ensure adequate coverage of different case managers and organizational settings. Data analysis uses a combination of inductive and deductive approaches to better understand the similarities and differentiation across implementation sites. Our findings add nuance to the contracting literature by elevating the implications of diversity and complexity in nonprofit program implementation.

Our study is part of a larger project evaluating the effectiveness of two different approaches to homelessness prevention: 1) a combined program of discretionary financial assistance and customized case management services, and 2) only flexible financial assistance. Using a randomized control trial, researchers at the University of Notre Dame’s Wilson Sheehan Lab for Economic Opportunities (LEO) are conducting the quantitative evaluation. Massachusetts Institute of Technology’s Abdul Latif Jameel Poverty Action Lab is funding both the quantitative and qualitative components of the research.