Association for Public Policy Analysis & Management

            State variation in implementation of Temporary Assistance for Needy Families (TANF), the program popularly termed “welfare” in the United States, has been widely studied. Existing scholarship, however, tends to focus on the rules, requirements, and activities associated with receipt of cash assistance. In addition to providing states with broad leeway over program design, the 1996 welfare reform law allowed federal and, in particular, state funds to be directed to other priorities consistent with the four goals of reform: assisting families in caring for children in their own homes, ending dependence on public supports through work and marriage, reducing the rate of out-of-wedlock pregnancy, and promoting two-parent family formation. Given this flexibility, cash assistance is no longer the primary use of TANF resources, accounting for less than one third of all program expenditures in 2013. States have instead used funds for such diverse activities as refundable tax credits, programs to prevent pregnancy and facilitate two-parent family formation, provision of childcare, and transfers to the Social Services Block Grant and Child Care and Development Fund Block Grant. Are the differing priorities of states under the umbrella of TANF related to differences in experiences of material hardship among low-income households with children? Addressing this question better aligns welfare scholarship with the contemporary nature of TANF as more than a cash assistance and employment program.

Methods

            Multi-level models including both state-level and household-level predictors of material hardship are estimated. The study merges two data sources: a dataset constructed from state TANF expenditure reports to the Administration for Children and Families and the annual Food Security Supplement to the Current Population Survey (CPS-FSS). The sample is restricted to low-income households, defined as a household below 185% of poverty, with at least one child present. The dependent variable, material hardship, is operationalized using CPS-FSS scales measuring household food insecurity. The key independent variables are state TANF expenditures devoted to various priorities, including basic assistance, work activities, alternative types of income transfer such as tax credits, and programs to prevent pregnancy and facilitate two-parent family formation. Second-level control variables, such as state unemployment rate to account for economic conditions, are also included as appropriate. Household-level variables related to hardship, such as race/ethnicity of the household head and highest level of education in the household, control for within-state variation. The study does not identify which households are actually participating in programs funded under TANF, relying instead on state expenditures as an indicator of overall approach to social welfare for low-income families. Despite this limitation, this inquiry better examines the relationship between alternative priorities, some of which target groups outside of the formal TANF program but for which there is a paucity of data, and the wellbeing of low-income families than do many existing studies of TANF.