Food Stamps and Welfare Reform

Michael Wiseman
Michael Wiseman Research Professor of Public Policy, Public Administration, and Economics - The George Washington University

March 4, 2002

In a typical month in 2001, 17.3 million people in 7.5 million households received food stamps at an annual cost of $20 billion. Current Food Stamp Program (FSP) appropriations expire September 30, 2002, coincidentally with the expiration of authorization for Temporary Assistance for Needy Families (TANF). The 1996 legislation that created TANF included food stamp provisions, and the close connections between TANF and the FSP mean that the welfare reform reauthorization debate involves both.

For low-income families with children, the FSP shares some characteristics with both TANF and the Earned Income Tax Credit (EITC). On the one hand, TANF dollars provide a general time-limited income floor for needy families that lack other means of support. On the other, the EITC bolsters the income of those who have jobs but work at low wages. The FSP supplements both TANF benefits and the incomes of the working poor and near poor while ensuring access to a necessity: food. And unlike the EITC, which is usually received annually as a lump sum, food stamp benefits are received monthly.

Evidence has accumulated over the past decade that the FSP is not functioning well as a support for working families. Some reforms have been implemented, and many others have been proposed. But before analyzing these reforms, this policy brief reviews the basic features and problems of the program.

How the Food Stamp Program Works

Food stamps help people buy food. Over 80 percent of recipients receive food stamp assistance by using special ATM-like debit cards in grocery stores. The basic benefit is adjusted annually for changes in food costs.

To be eligible for food stamps, households must have gross (before tax) monthly incomes of less than 130 percent of federal poverty guidelines (in 2001 the guideline was $1,219 for a family of three) and few assets. Certain adults are required to register for work, and some adults without dependents are required to work or to participate in training as a condition of assistance. Families receiving TANF benefits and persons receiving Supplemental Security Income are in most circumstances automatically eligible for food stamps if they live alone.

In fiscal year 2001, the maximum monthly food stamp benefit for a household of three was $341. Beyond a standard deduction and certain other allowances, benefits are reduced by $0.30 for each dollar of income from sources other than earnings and by $0.24 for each dollar of earnings. In fiscal year 2000, households with children received an average monthly food stamp benefit of $234.

While the federal government pays most FSP costs and sets most of the regulations, the program is operated by states, generally through local welfare offices. Management is evaluated annually by a joint federal/state review of a sample of cases drawn from each state’s recipients list. This quality control sample is sufficiently large to provide reliable information on the people receiving food stamps, the rate at which administrators make errors in benefit determination, and the amounts of payments involved. States can be charged for the benefit cost of error rates in excess of national averages. In practice such penalties are often waived; when enforced, states pay by investing the fine in programs to improve performance.

Food Stamps Play Two Roles

Food stamps both help protect household access to a necessity and provide income support. The safety role is reflected in the program’s focus on monthly income and on adjusting benefits relatively quickly when income changes. Beyond the safety net, food stamps help fill long-term gaps between the income people need, at least for nutrition, and what they have. The food stamp benefit serves as a supplement to a family’s other resources, usually Supplemental Security Income, TANF, or earnings. In at least fourteen states, TANF families receive more each month in food stamps than in TANF cash.

In its income support role, the FSP is an important contributor to the well-being of the working poor and near poor who are not TANF recipients. In 2001, a single mother with two children who worked thirty hours a week at $8 an hour earned about $950 a month after social insurance deductions—less than the poverty standard of $1,219 for a family this size. But when she files her federal income tax return, she will get almost $4,000 ($324 per month worked) from the EITC. In addition, she is eligible each month for at least $134 in food stamps. While for most working households the EITC benefit is greater than the food stamps benefit, EITC income is usually not available to meet ongoing expenses until the end of the year.

From the perspective of welfare policy, TANF, the FSP, and the EITC should work together: TANF gives states program flexibility for addressing general family needs and moving people into the workforce; the FSP provides a real-time income supplement; and at tax time, the EITC provides a special cash bonus for families working at low wages and contributing to Social Security. Combined, these programs bring most low-wage families that work full-time above the poverty level. Thus, the idea of food stamps as a bridge from welfare to work is attractive. But recent data suggest that the bridge may not be working as well as it should.

The Big Decline in Participation

As illustrated in figure 1, between 1993 and 2000 the number of persons counted as poor fell by 21 percent, from 39 million to 31 million. The number of children living in poor households declined as well, by 26 percent, to 12 million. These trends would normally be expected to reduce food stamps usage. But especially after 1995, the number of FSP participants fell by considerably more than would seem warranted by the observed reduction in poverty. Since most poor families are eligible for food stamps, the implication is that the percentage of eligible families that actually received the benefit fell.

It is difficult to estimate take-up rates with precision, because available household surveys do not provide the detail on assets and monthly income necessary to simulate reliably FSP eligibility determination procedures. Nevertheless, the best available estimate is that food stamp take-up fell by 24 percent between 1993 and 1999, and the decline accelerated following welfare reform. Consequently, by 1999 as many as 4 million children were living in households eligible for, but not receiving, food stamp assistance. The decline appears particularly pronounced among both single-parent and married-couple families with earnings. Given that a major objective of welfare reform is to encourage work and to promote marriage, these changes are of special concern

Why has this occurred? Many analysts point to the administrative tension between TANF and food stamps as a factor in at least the early decline. Food stamps are a federal entitlement, while the 1996 reforms made eligibility for TANF a matter principally of state policy. States often attempt to avoid establishing new cases for TANF applicants by meeting particular one-time needs or by rapid job placement. Given such “diversion” strategies, application for food stamps, if not discouraged, was often not promoted. In response to studies by the U.S. General Accounting Office and efforts by both the Food and Nutrition Service of the Department of Agriculture (which runs the FSP) and by advocates, states eventually moved to clarify the distinction between food stamps and TANF and to make more information about food stamps available to all persons seeking assistance.

Administrative problems also arise in the case of families leaving welfare. It appears that many families leaving TANF have been unaware of the possibility of continued eligibility for food stamps assistance. States have responded with efforts to better inform families about the FSP, and states may now allow food stamp benefits to continue automatically for three months after TANF benefits cease in order to facilitate transition.

The last decade saw a substantial increase in employment among low-income families, especially among single parents. Somewhat paradoxically, this desirable development may have contributed to the decline in food stamp take-up. More earnings means lower food stamp benefits. At the same time, working takes time, making it more costly to come into welfare offices to review the food stamps application. This apparently leads some families to give up food stamps even when still eligible for significant benefits.

In some instances state policy may have added to the difficulty families encounter in trying to sustain food stamps eligibility. Working households generally have more variable incomes than do households dependent on other sources of income. Accordingly, benefit computation errors tend to be more common for working families. States have apparently responded to this hazard by increasing food stamp reporting requirements for workers, thereby raising the “hassle factor” even further for households facing the greatest time constraints. To offset this development, the Department of Agriculture has changed program regulations so that states can reduce the reporting burden without generating errors that will lead to federal penalty.

There is some evidence that efforts to streamline the program since 1996 and more aggressive outreach efforts have increased utilization: participation began to rise in early 2001. However, the near-simultaneous economic slow-down confounds the interpretation of such changes. What is clear is that making work pay is central to national welfare strategy, and if, for whatever reason, eligible working people lose access to food stamps, reform belongs on the agenda for reauthorization.

Policy Opportunities

The core objective of the FSP is to provide nutrition support for people in need. The central question for reform is how to find efficient ways to improve access to the program’s benefits for those it is intended to serve without compromising other national objectives such as fraud reduction and promotion of self-support.

Working Families

For working families, the options for food stamp reform most commonly cited involve changing rules and administrative procedures. More ambitious possibilities include changing the way the benefit is delivered

It is possible that estimates of low FSP take-up are exaggerated because families qualified for food stamps on the basis of income do not apply, or are determined ineligible, because of their assets. The liquid assets maximum ($2,000) was set in 1985; the maximum automobile value ($4,650) was set in 1996. The assets maximum could be increased by over 60 percent without exceeding amounts equivalent in purchasing power to those used in the past; the automobile maximum is obviously too low for households that need a reliable vehicle for commuting. Under regulations established by the Department of Agriculture in 2000, states are now allowed to align the FSP vehicle maximum to the limits in their TANF programs—if these limits are higher—and a majority of states have done so. Updating and increasing both the vehicle and liquid asset limits nationally would be consistent with other efforts to promote saving and self-sufficiency, since households that save create their own safety nets. Current proposals in both the House and Senate farm bills generally leave vehicle policy to the states; the administration’s proposal calls for exempting one vehicle per adult.

Turning to procedural issues, the FSP requires regular “recertification” of individual and family need. In most jurisdictions, recertification is done at quarterly intervals for families with earnings, which is a burden for working families. The obvious solution is to fix benefit values over a longer period of time. Regulatory changes made by the Department of Agriculture in late 2000 give states the option of setting benefits semi-annually for earners while allowing households to apply for a benefit adjustment if income falls. Senate farm bill proposals call for extending this option to virtually the entire caseload.

Frequent recertification is costly, both for agencies and for clients, especially those clients who are working. The problem is one of conflict between the requirements of a safety net program—meeting the nutrition needs of the moment—and the longer-term perspective appropriate for supplementing the low, and sometimes unsteady, incomes of working families. A system that fixes benefits for an extended period on the basis of current income increases the gain from applying for food stamps when income is low, even when recovery is imminent. In principle, this issue might be addressed by reconciliation of benefits with actual income at, say, income-tax time. But to do so would require more elaborate income reporting than is now done and possibly integrating the food stamps program with the income-tax reporting system. Given such integration, consideration might be given to converting the food stamp benefit to cash, at least for working families.

But taking this step of converting food stamps to cash would seem to contradict the fundamental orientation of the FSP toward nutrition. Evidence exists that providing aid in the form of food stamps does in some cases lead households to purchase more food than they would if the benefit were provided in cash. However, this does not mean that the Electronic Benefit Transfer (EBT) card, which is now used by most states, is the best way of distributing FSP benefits. Payment through EBT is most likely to effect consumption for those with the lowest incomes-the families for whom the program’s safety net function is most important. For those families with benefits that are small in comparison with overall expenditures—typically families with earnings—the FSP principally provides the equivalent of more income. In such cases both the nutrition and the income support function might be served equally well by depositing the benefit in a bank account. Without experimentation, it is not clear how much this shift would raise FSP participation.

Some food stamp applicants and recipients do cheat. Suppressing fraud is important to maintaining the FSP’s political viability, and regular assessments of eligibility and review with recipients of their obligations contribute to the program’s integrity. Nevertheless, the gains from using an expanded horizon for eligibility assessment and simplified processes for recertification appear substantial. Rather than introduce such changes by directive, as the Department of Agriculture has done for some features, it might be better to undertake careful experimentation in order to determine the best among a number of options for modification of income assessment and reporting procedures. Current food stamp proposals in both the House and Senate include provisions for funding of demonstrations intended to test alternatives for FSP regulation and operation

Integration with TANF

Problems with integration of food stamps with TANF arise in three ways. One, mostly resolved in the 1996 welfare reform law, concerns sanctions. When adult TANF recipients lose benefits due to failure to comply with TANF rules, their income falls. Without adjustment, a fall in income would produce an increase in the food stamps benefit, offsetting the sanction. Current regulations allow states to preempt such adjustment for the adult portion of the food stamp benefit.

A second TANF integration issue involves assets. States have in many instances adopted more generous assets restrictions for TANF than for food stamps. This does not pose a problem for TANF recipients, since TANF receipt automatically qualifies families for food stamps. However, it is possible for a working family to leave assistance and lose food stamps not because of application of the income test but because of more stringent FSP vehicle or savings restrictions. This inconsistency has been addressed by giving states a variety of options for relaxing federal rules. The result is a hodgepodge.

The pressure to adapt FSP rules to state TANF choices has occurred in part because of failure to adjust the federal food stamp limits upward. If federal policy were consistent with the support role the FSP is expected to play for working families, it would be reasonable to expect states to adapt their TANF policies to the federal program. But when the federal government is not actively involved in program review and improvement, it is not surprising that pressure mounts for accommodating national policy to the varied strategies, wise and otherwise, of the states.

A third integration issue is related to closure of TANF cases. It is awkward and possibly counter-productive to immediately conduct a new eligibility determination for food stamps as families are leaving TANF. Current regulations allow states to treat the value of the benefit at closure as transitional assistance, and to sustain the payment for three months before undertaking recertification. This transition could be lengthened: extension to six months would be consistent with current practice in Medicaid and could be linked to a more general policy of extended benefit determination like that outlined earlier. Food stamp proposals in both the House and Senate include provisions for extending the transition benefit, although details differ.

Federal Oversight

Because states bear a substantial share of FSP administrative costs but pay nothing for benefits, fiscal incentives for careful administration are diminished. States contribute to the increased management costs often required for raising efficiency and participation, but they gain nothing from savings generated by reducing inappropriate payments or successes achieved in raising participation. The quality control system, which imposes careful checks on state administration, is the federal government’s attempt to deal with this dilemma. There is little doubt that states improve their performance when problems are revealed by quality control assessments. But auditors focus on rules, and rules make systems inflexible.

Proposals for change tend to be of four types: to eliminate or weaken the quality control system, to change what is considered an error, to change the state reward structure, or to assist states with error rate reduction. The desire to eliminate the quality control system is understandable; the FSP is one of the few government programs where comparative performance by states is assessed and publicized. Nevertheless, there are other options for addressing shortcomings of the program.

State liability for errors is currently based on comparison of state error rates to national averages, something that can be known only after the fact. An alternative procedure, helpful to state planners, would be to fix national targets for error rates and to assess rewards and penalties relative to these rates. Such target rates could be adjusted to reflect special circumstances. In 2000, the Department of Agriculture introduced adjustments in its liability assessment procedures to offset the potential cost to states of errors generated by increases in working family FSP participation. However, instead of adjusting fiscal liability to reflect state caseload or population characteristics, it would be better to adjust the targets themselves. Current House and Senate farm bills call for adjusting procedures for error determination and penalty assessment, but retain focus on average error rates.

Accuracy in food stamps benefit determination is surely not the only objective of national food stamps policy, yet current quality procedures focus virtually all attention on such considerations. Recent changes in recertification requirements and other administrative procedures clearly promote access by making outreach to working families less risky. But if access is important, performance assessment should be expanded beyond issues of computation and documentation to include, for example, measures of state provision of opportunity for food stamp application and recertification when and where they are most useful to working families.


There are three overarching issues in food stamps policy discussions: defining what the program is intended to do, structuring the relationship between food stamps and cash welfare, and defining the terms of the federal-state partnership required for achieving food stamps goals. Reauthorization presents an opportunity for progress in all three areas. What seems most important is a new approach to the partnership itself, one that recognizes the unique national character of the Food Stamp Program but seeks ways to more systematically integrate state experience and capacity in the process of national program improvement and evolution. Thinking this way might move the debate from reauthorization to revitalization.