Interrelationships among public assistance expenditures: An empirical analysis of the welfare system

This article uses vector autoregression techniques with a panel data set to test for interdependencies between spending on the Aid to Families With Dependent Children (AFDC) program and expenditures on three other major public assistance programs: Medicaid,food stamps, and Supplemental Security Income. Tests of Granger causality reveal that there ape spillover effects in both directions among these programs, suggesting that substantial feedback effects likely. The author uses the estimated model to simulate the effects of one-time spending decreases an either side. The simulation, which includes both direct and indirect effects, shows that these spending cuts results in long-term decreases in expenditures an the affected program. However changes in AFDC spending are likely to be at least partially offset by changes in one or move of the other programs. Similarly, budget cats in the other programs can result in increased AFDC spending. This suggests that the new welfare law may not produce the budgetary savings expected by its sponsors.