Methods: We use a newly expanded comprehensive Wisconsin administrative research dataset that includes records for the entire state population across multiple programs and over time. We follow three different populations: participants in Temporary Assistance for Needy Families (TANF), Supplemental Nutrition Assistance Program (SNAP), and Unemployment Insurance (UI) in Wisconsin in 2005, 2007 and 2009. We take into account multiple program participation and disconnection in the sample year as well as in the previous and subsequent years. Resources from TANF, SNAP and UI benefits, as well as wages from UI covered employment, Supplemental Security Income (SSI) benefits, and child support are considered. We also measure participation in subsidized child care and Medical Assistance (MA), interactions with the child welfare system, and incarceration in the State prison and county jail systems. We document the incidence and prevalence of disconnection and how this disconnection varies over time and across populations defined by previous employment and program participation, region, and demographic characteristics.
Results: We find there are differences in incidence and pattern of multiple program participation across TANF, SNAP and UI recipients. Although a substantial number of recipients within TANF, SNAP and UI leave these specific programs in the subsequent years, they are found to circulate and participate in other income support programs. For instance, there are substantial increases in participation in SNAP and MA among former UI beneficiaries; and increases in SSI and Child Support income among TANF and SNAP beneficiaries. We also find substantial numbers of TANF, SNAP and UI beneficiaries who appear to be disconnected from all or most public benefits and the formal labor market in the subsequent years. The risk of the disconnection varies by demographic characteristics, region, and family circumstances. Using data through 2010, we also find evidence of substantial changes in program participation over the business cycle, as less disadvantaged people participate more intensively in services in the recent recession.
Concludsion and Implications: Our results highlight the population at greatest risk of becoming disconnected, informing program and policy responses. By analyzing disconnection as one end of a continuum of potential multiple program participation we are better able to identify the factors that explain participation dynamics. It is imperative to gain a better understanding of the challenges that the disconnected population faces in order to provide better and more efficiently targeted services.