Association for Public Policy Analysis & Management

Temporary Assistance for Needy Families (TANF) is often viewed as a cash assistance program with accompanying work activities and services for low-income families. TANF’s policy design—a federal block grant with a state maintenance-of-effort (MOE) contribution—provides states with extensive flexibility over the use of program resources, however. It also includes incentives (e.g., caseload reduction credits) for states to use federal and MOE funds for purposes other than traditional cash benefits. Basic assistance now comprises roughly one quarter of all TANF spending; even combining basic assistance with work supports and childcare still accounts for only approximately half of all block grant and MOE funds. Rather than a cash assistance program, TANF is best considered as a funding stream of which cash assistance is one possible use. Most studies of state TANF implementation, though, examine predictors of cash assistance rules and requirements.

States vary considerably in their uses of TANF funds, and there is a wide range both in emphasis on cash assistance (in 2013, cash benefits accounted for 52% of Maine’s TANF allocations but only 7% of Illinois’s) and rate of decline in cash assistance spending over time (ranging from a 93% decrease from 1998 to 2013 in Illinois to a 3% increase in Virginia). In this study, I address two questions bridging the gap between the current form of TANF and scholarship on state TANF politics. First, what explains variation in state use of TANF funds for basic assistance? Second, what explains state differences in the decline in cash assistance over time? I pay particular attention to two factors. First, race-based politics is related to other aspects of state cash assistance policy (e.g., states with larger proportions of people of color on the cash assistance caseload adopted stricter rules). TANF affords states in which race is salient to policymaking the opportunity to de-emphasize cash assistance. Second, I consider the influence of state finances on TANF expenditures. States under fiscal stress could divert funds from basic assistance to other priorities, limiting countercyclical responsiveness.

Methods

I estimate multilevel growth curve models of state TANF basic assistance expenditures from 1998 to 2013. I express the basic assistance expenditures dependent variable in two ways, scaled to the estimated count of families in poverty in the state and as a percentage of total TANF spending. In contrast to most previous studies, I operationalize the salience of race to state politics with respect to blacks using a measure of the prevalence of extremely negative stereotypes of blacks among whites in the state. This variable is constructed using multi-level regression with post-stratification, a technique for producing subnational public opinion estimates from national data, with the 2008 National Annenberg Election Survey dataset. Fiscal stress is measured using the difference between state expenditures and state revenue scaled to account for variation in the size of state budgets. I also control for other relevant political, social, and economic factors, including government political ideology, citizen political ideology, unemployment rate, and unmarried birthrate.