Association for Public Policy Analysis & Management

Our approach is to describe sub-populations of SNAP households using latent class analysis.  Identifying the characteristics and sizes of these sub-populations, or classes, is key for helping policymakers make resource allocation issues, but also to understand the condition of the poor in a state and how that changes across the economic cycle.

In consultation with Illinois policymakers, we found that they preferred an analysis of yearly point-in-time caseloads to understand how their task of providing work supports changed over time and how it may be driven by the characteristics of the caseloads.  Therefore, for each caseload, we looked both retrospectively and prospectively to understand prior and future status of the case—as a function of the work supports.  This allows us to provide a three-year window for each annual caseload and provides policymakers with a way to think about how to evaluate their prior use of resources, as well as plan future spending and allocations. A portion of this study has already impacted the child care subsidy program by the state deciding to broaden the program after shrinking it to only TANF recipients in 2015.

The classes that emerge are intuitive, but the sizes are somewhat surprising.  We see a small portion of zero-earned income cases, while we see a fairly large percentage of these families with young children using child care subsidies.  Some of these increase their income and some don’t, which creates two classes.