Consumer Debt and Program Participation

In this paper, we examine how US households' access to credit affects their participation in safety-net programs. We examine how participation in the Food Stamp Program (FSP)/Supplemental Nutrition Assistance Program (SNAP) and the Temporary Assistance to Needy Families (TANF) program respond to changes in state level debt relative to income using 20 years of quarterly level data, from 1999-2019. Simply modeling FSP/SNAP participation, TANF participation, and debt-to-income as independent dynamic panel models, we find that an increase in debt-to-income levels has, at best, a modest and weakly positive effect on FSP/SNAP and TANF participation. However, we argue this fails to capture the full systemic structure of the safety-net and household finances. We employ a panel vector autoregression model to capture debt relative to income and program participation as a system in a causal framework. We show that TANF is unresponsive to a debt-to-income shock, while FSP/SNAP participation significantly, substantially, and persistently increases following a positive debt-to-income shock, rising by nearly 1.4% cumulatively over five years following a moderate 1 percentage point shock to the debt-to-income ratio. We hypothesize this is due to the differing nature of the programs, and an important consideration when designing safety-net policy