State Budgetary Behavior and the Fiscal Interactions Between Food Stamps, AFDC, Medicaid, and SSI

Year: 1999

Research Center: Joint Center for Poverty Research, University of Chicago and Northwestern University

Investigator: Chernick, Howard

Institution: Hunter College, City University of New York

Project Contact:
Howard Chernick, Department of Economics
Hunter College, City University of New York
695 Park Ave.
New York, NY 10021
212-772-5440, fax 212-772-5398


The Federal Government has full control over the eligibility requirements and benefit structure for the Food Stamp Program, which is financed mainly with Federal funds. By contrast, the States have substantial powers to set the requirements for several cash assistance programs, and they must pay a substantial portion of the costs of these programs out of State funds. By controlling the rules for cash programs, States can influence the amount of all types of benefits each household receives—including food stamp benefits—and at the same time influence total State spending on welfare programs. In this study, Chernick develops a model to measure the extent to which States may be using their influence to shift more of the costs of welfare to the Federal Government by substituting food stamps for cash assistance. Depending on the preferences and budgeting rules used by States, funds released by substituting food stamps for cash assistance can be used to fund other welfare programs, provide tax relief, or to supplement other State and local spending.

Most recipients of Aid to Families with Dependent Children (AFDC)/Temporary Assistance for Needy Families (TANF), Supplemental Security Insurance or General Assistance are also eligible for food stamps.

The link between cash assistance and food stamps is the way cash income is counted in determining food stamp benefits. Food stamp benefits are based on a household’s “net income” after certain deductions, including a standard deduction and a deduction for some housing expenses. A household with no “net income” receives the maximum food stamp benefit. Each dollar of positive “net income” causes a reduction of 30 cents in food stamp benefits. In this sense, the State’s contributions to cash assistance programs are implicitly “taxed” at a rate of 30 percent. The actual values of this implicit “tax” in each State vary for different values of the food stamp shelter deduction.

To estimate the extent of substitution of food stamps for cash assistance, Chernick used variation over time in the food stamp maximum benefit, the standard income deduction, the maximum excess shelter deduction, and variation among States in the average “tax” rate on cash benefits. He estimated the effect of these variables on AFDC benefits per recipient, total AFDC spending per capita, and Medicaid spending. Data are from the 48 contiguous States and the District of Columbia for the period 1983-95.

The interaction between the implicit “tax” and the excess shelter deduction results in actual implicit “tax” rates ranging from 30 percent to 45 percent. The average cost to States of raising the income of AFDC/TANF recipients by $1.00 is a function of this “tax” rate and the proportions of the State’s food stamp recipients receiving a shelter cost deduction and recipients above the shelter cap. The average cost of $1.60 is quite high, and potentially serves as a strong deterrent to States contemplating raising their benefit levels.

Chernick’s preliminary results suggest a significant and economically large effect of the food stamp “tax” on cash assistance. The estimated effects on cash benefits are large enough to imply that a decrease in the food stamp implicit “tax” on AFDC/TANF benefits would lead to both an increase in cash benefits and a decline in food stamp outlays. However, he emphasizes that even with the excess shelter deduction taken into account, there is not enough variation in the implicit “tax” that States face to be very confident of the results. Hence, at this stage of the research, the estimates must be viewed as highly tentative. The results also show an almost dollar-for-dollar offset of food stamps for cash resulting from increases in the food stamp maximum benefit and in the deductions from income.

Chernick also found evidence that funds saved by substituting food stamps for cash assistance were used to increase Medicaid spending on AFDC recipients, suggesting that at least some of the States’ savings remain within their welfare budgets.

Chernick cautions against using the estimates for forecasting effects of changes in food stamp implicit “tax” rates outside of the sample range. Because the analysis is based on program rules set prior to welfare reform, it is not appropriate to apply the results to the post-welfare-reform environment. He is conducting further research using alternative instruments to try to improve the accuracy of the estimates.