Reforming the Food Stamp Program

Report Hunger and Food Programs

Reforming the Food Stamp Program

July 25, 2012 29 min read Download Report

Authors: Katherine Bradley and Robert Rector

Abstract: The food stamp program is due for reauthorization as part of a new farm bill. It is the second most expensive means-tested aid program, increasing from $19.8 billon in 2000 to $84.6 billion in 2011, and President Barack Obama has proposed a budget to keep food stamp spending at sharply elevated levels for the next decade. The national debt has topped $16 trillion and will continue to grow rapidly for the foreseeable future. To preserve the economy, government spending, including welfare spending, must be put on a more prudent course. Congress and the Administration should transform food stamps into a program that encourages work and self-sufficiency, close eligibility loopholes, and, after the recession ends, reduce food stamp spending to pre-recession levels.

The farm bill is due for reauthorization, including the food stamp program, also known as the Supplemental Nutrition Assistance Program (SNAP). The food stamp program is very large and growing rapidly. President Barack Obama plans to spend nearly $800 billion on food stamps over the next decade.[1] Yet as large as it is, this program is only part of a much larger system of means-tested government assistance: 79 programs that provide cash, food, housing, medical care, and social services to poor and low-income Americans. President Obama’s fiscal year (FY) 2013 budget calls for $12.7 trillion in means-tested aid over the next 10 years.[2]

Means-tested welfare is the fastest growing component of government. Total federal and state means-tested spending—which excludes Social Security, Medicare, and Unemployment Insurance—rose from $431 billon in 2000 to $927 billion in 2011. According to President Obama’s spending plans, annual means-tested spending will rise to $1.2 trillion within four years and $1.5 trillion per year by 2022.[3]

The food stamp program is the second most expensive means-tested aid program. Food stamp spending has grown rapidly in recent years, from $19.8 billion in 2000 to $84.6 billion in 2011. Part of that growth is due to the recession, but under Obama’s proposed budget, food stamp spending will not return to pre-recession levels when the economy recovers. Instead, it will remain well above historic norms for the foreseeable future.

The food stamp program is old and fossilized. Aside from enormous increases in cost, it has remained basically unchanged since its creation in the 1960s. Unaffected by welfare reform in the 1990s, it remains a program that discourages work, rewards idleness, and promotes long-term dependence.

The national debt has now topped $16 trillion and will continue to grow rapidly for the foreseeable future. To preserve the economy, government spending, including welfare spending, must be put on a more prudent course. When the current recession ends, food stamp spending should be returned to pre-recession levels. In addition, Congress should transform the program from one that rewards dependence into one that encourages work and self-sufficiency.

To accomplish, this Congress should:

  1. Return food stamp spending to pre-recession levels and cap future spending. The food stamp program is an open-ended entitlement. State governments receive automatic increases in food stamp funding when they increase the number of recipients on the food stamp rolls. This practice encourages high levels of spending and unnecessary dependence on government aid. In the future, the uncapped entitlement nature of spending on food stamps should be ended. When the current recession ends, Congress should return aggregate spending to pre-recession levels. In subsequent years, the maximum allocation to states should grow no faster than inflation and population growth, although temporary increases above that maximum could be permitted in periods of high unemployment.
  2. Transfer control over food stamps from the Department of Agriculture (USDA) to the Department of Health and Human Services (HHS) . The food stamp program is a means-tested welfare or anti-poverty program, not an agricultural program. The USDA’s expertise is in farming, not welfare; the department has shown that it is unsuited to run the second largest means-tested aid program in the nation. Authority over food stamps should therefore be transferred from the USDA to HHS.
  3. Eliminate application loopholes that permit food stamp recipients to bypass income and asset tests. These loopholes artificially inflate caseloads and costs.
  4. Reduce fraud. Unlawful benefit overpayments should be reduced by acquiring more timely and accurate information about recipient earnings.
  5. Prohibit food stamp payments to illegal immigrant families. Illegal immigrants who have children born in the United States routinely receive food stamps. This policy should be changed and brought in line with existing policy on the earned income tax credit (EITC). In order to receive the EITC, parents must provide a valid Social Security number demonstrating they are residing in the U.S. lawfully and are authorized to work. The same policy should be applied in food stamps.
  6. Convert food stamps into a work activation program. Able-bodied food stamp recipients should be required to work, prepare for work, or at least look for a job as a condition of receiving aid. These work activation requirements should be phased in gradually as the current economic recession ends. Similar requirements in the Aid to Families with Dependent Children (AFDC) program in the 1990s led to dramatic drops in welfare caseloads, surges in employment, and a large-scale drop in child poverty among groups that were most dependent on the program.
  7. Require drug testing of food stamp recipients. As the federal government faces future bankruptcy, scarce taxpayer funds should not be used to provide welfare to individuals who abuse illicit drugs. In the food stamp program, taxpayers should not be required to pay for free food for individuals who waste their own money on illegal drugs. Food stamp applicants and recipients should be tested for illegal drug use, and benefits should be terminated for those using drugs. Drug testing would reduce wasteful government spending. It would also reduce drug use among food stamp recipients, thereby increasing their future employability.

These reforms are generally modeled on the 1990s welfare reform, which replaced the AFDC program with the Temporary Assistance to Needy Families (TANF) program. That reform slowed the growth of welfare spending and increased employment while reducing both dependence and child poverty. It enjoyed widespread public support.

Understanding the Means-Tested Welfare System

The food stamp program is very large and growing rapidly, but it is only part of a much larger system of government means-tested assistance. For example, SNAP is only one of 12 federal programs that provide food aid to the poor.

Most people who receive food stamps also participate in other government aid programs. It is therefore misleading to examine food stamps in isolation. Such an approach inevitably underestimates the level and cost of assistance provided to the poor. Food stamps should be analyzed holistically as one component of a much larger means-tested welfare system of 79 federal programs providing cash, food, housing, medical care, social services, training, and targeted education aid to poor and low-income Americans.

Means-tested welfare programs differ from general government programs in two ways. First, they provide aid exclusively to persons (or communities) with low incomes. Second, individuals do not need to earn eligibility for benefits through prior fiscal contributions. Means-tested welfare therefore does not include Social Security, Medicare, Unemployment Insurance, or Worker’s Compensation.

In FY 2011, the federal government spent $717 billion on means-tested welfare. State contributions to federal programs added another $201 billion, and independent state programs contributed around $9 billion. Total spending from all sources reached $927 billion.[4]

Means-tested spending amounts to $9,040 for each lower-income American (i.e., persons in the lowest third of the population by income). If converted to cash, this spending is more than sufficient to bring the income of every lower-income American to 200 percent of the federal poverty level (roughly $44,000 per year for a family of four).[5]

In the two decades before the current recession, means-tested welfare was the fastest growing component of government spending. It grew more rapidly than Social Security and Medicare, and its rate of increase dwarfed the growth of public education and national defense. While means-tested medical benefits have been the fastest growing part of the welfare system, most other forms of welfare aid also have grown rapidly.

According to the President’s FY 2013 budget, means-tested welfare will continue to grow rapidly for the next decade instead of declining as the recession ends. The President’s budget would permanently increase annual means-tested spending from 4.5 percent to 6 percent of gross domestic product. Combined annual federal and state spending would reach $1.56 trillion in 2022. Overall, President Obama plans to spend $12.7 trillion on means-tested welfare over the next decade.

The President’s budget calls for ruinous and unsustainable budget deficits. These deficits are in part the result of dramatic, permanent increases in means-tested welfare. An important step in reducing future unsustainable federal deficits would be to return total welfare spending to pre-recession levels.

To accomplish this, Congress should cap future aggregate welfare spending. When the current recession ends, or by 2013 at the latest, total federal means-tested welfare spending should be returned to pre-recession levels, adjusted for inflation. In subsequent years, aggregate federal welfare spending should grow no faster than inflation. This type of spending cap would save the taxpayers $2.7 trillion during its first decade.

An aggregate welfare spending cap of this sort is contained in the Welfare Reform Act of 2011 (H.R. 1167), introduced by Representative Jim Jordan (R–OH), and a companion bill (S. 1904), introduced by Senator Jim DeMint (R–SC).

1. Capping Future Food Stamp Spending

Just as it is critical to restrain the rapid growth of overall means-tested spending, it is also important to limit excessive spending in the food stamp program individually. The federal government pays the full cost of food stamp benefits and splits administrative costs with state governments that administer the program. In FY 2011, federal spending was $77.6 billion, and state costs were approximately $6.9 billion.

As noted, the food stamp program is growing rapidly. Before the current recession, combined federal and state spending nearly doubled, rising from $19.8 billion in 2000 to $37.9 billion in 2007. Since taking office, the Obama Administration has more than doubled spending on food stamps again: Spending rose from $39 billion in 2008 to a projected $85 billion in 2012. (See Chart 1.) Even after adjusting for inflation and population growth, food stamp spending is now nearly twice the level in any previous recession.

The current recession has obviously caused part of the overall spending increase, but the USDA’s Food and Nutrition Service has also liberalized eligibility standards and operated aggressive outreach programs for more than a decade with the goal of maximizing the number of food stamp recipients. These efforts, combined with the recession, have swollen the food stamp caseload to well above normal historical levels.

Moreover, President Obama’s FY 2013 budget shows that the President does not intend food stamp spending to return to pre-recessionary levels. Instead, outlays will remain at historically high levels for the foreseeable future. For most of the next decade, food stamp spending, adjusted for inflation and population growth, would remain at nearly twice the levels seen during the non-recessionary periods under President Bill Clinton.

This long-term increase in food stamp spending is not sustainable. In keeping with the general aim of controlling the overall rapid growth of means-tested welfare, Congress should reduce the abnormally high levels of future food stamp spending by taking the following steps.

  1. After the current recession, Congress should return total federal spending on food stamps to pre-recession levels adjusted for population growth and inflation.[6]
  2. In subsequent years, food stamp spending should grow no faster than the rate of inflation combined with population growth.
  3. During periods of very high unemployment, spending may temporarily exceed this limit.
  4. Congress should provide each state with an annual food stamp allocation based on its pre-recession spending level adjusted for inflation and population growth.

To implement this cap, the entitlement nature of food stamp spending should be eliminated.[7] Automatic open-ended increases in spending should be curtailed, and states should be given greater flexibility to determine program eligibility. A food stamp spending cap of the sort described above would save the federal government roughly $150 billion over the next decade.

Overall, the government should make an effort to return food stamp caseloads to normal, pre-recession levels or to the even lower levels experienced during the Clinton presidency. The additional reforms described below would contribute to that process.

2. Transfer food stamps from the USDA to HHS.

The purpose of the U.S. Department of Agriculture is to assist farming. The USDA’s expertise is in agriculture, not welfare. But the food stamp program and similar means-tested assistance programs such as the Women, Infants, and Children (WIC) food program and school lunch and breakfast programs are now two-thirds of the USDA budget. The USDA has become primarily a welfare agency with some farm programs tacked on the side. It is ill suited for this task.

Historically, the USDA has viewed food stamps and other means-tested food aid programs as a mechanism to expand farmers’ income by promoting increased food purchases.[8] This is an inefficient policy to provide income to farmers; for each $1.00 government spends on food stamps, farmers receive about three cents in added income.[9] It is also a poor foundation on which to build welfare and anti-poverty strategies. Given its background, it should be no surprise that the food stamp program grows rapidly and is highly resistant to the main ideas of welfare reform.

Food stamps should be reformed along the lines of other modern welfare programs like TANF, not continued as a farmers’ aid program. Its aim should be to provide support to low-income households while encouraging work and self-sufficiency. To accomplish this, authority over food stamps and related USDA means-tested aid programs should by transferred from the USDA to HHS. Reauthorization of the food stamp program should be accomplished as stand-alone legislation; it should not be part of the largely unrelated farm bill.

3. Close expensive loopholes in food stamp enrollments.

The food stamp rolls have skyrocketed, in part because the Obama Administration has promoted admission procedures that enable applicants to bypass ordinary income and asset tests for eligibility. One such policy is known as “broad-based categorical eligibility.” Within the welfare system, a single family may often be eligible for several welfare programs. As a way to avoid duplicative administrative costs, receipt of aid from one program may give categorical eligibility for another.

Historically, categorical eligibility was intended to simplify program administration and did not greatly expand food stamp eligibility; however, in 2000, the Clinton Administration quietly issued regulations concerning the relationship of food stamps to the TANF program, creating a radical new food stamp admission procedure termed “broad-based categorical eligibility.” This procedure dramatically altered the rules of food stamp eligibility.[10]

Traditionally, categorical eligibility in food stamps pertained only to individuals receiving cash assistance from selected other programs; in most cases, eligibility for these other programs was narrowly defined by the federal government. The 2000 regulation gave states the option to provide broad-based categorical eligibility for food stamps to any person or household that received cash aid or any other service funded by TANF.[11]

This was a major change because, under the TANF program, states are given broad discretion in determining who will receive TANF cash aid.[12] In addition, most people in a state are potentially eligible for TANF non-cash services. Since virtually anyone can be eligible for a TANF non-cash service irrespective of income or assets, this rule gave states the option to substantially broaden their food stamp eligibility standards, weakening the income limits and waiving the asset limits entirely.[13] States could make this change without incurring added fiscal costs on themselves.

Historically, the food stamp program was limited to persons with both low incomes and limited liquid assets. Individuals with more than $2,000 in liquid assets could not receive assistance; households were expected to use their own assets to support themselves before turning to taxpayer-funded welfare.[14] After 2000, an increasing number of states chose to use the new broad-based categorical eligibility loophole to eliminate asset limits for food stamp recipients and greatly expand their food stamp caseloads.

The Obama Administration has actively promoted use of the broad-based “categorical eligibility” loophole by states. According to the Congressional Research Service:

The USDA’s Food and Nutrition Service has taken an official stance encouraging states to use so-called “categorical eligibility” authority to expand eligibility to significant numbers of households by (1) increasing or completely lifting limits on assets that eligible households may have and (2) raising dollar limits on households’ gross monthly income.[15]

By 2012, 43 states used receipt of a nominal “service” from TANF to establish broad-based categorical eligibility for food stamps.[16] In these states, the food stamp application process works as follows: An individual applying for food stamps is automatically handed a brochure printed with TANF funds. The Congressional Research Service describes the process as follows:

Typically, households are made categorically eligible [for SNAP benefits] through receiving or being authorized to receive a minimal TANF- or MOE-funded benefit or service, such as being given a brochure or being referred to a social services “800” telephone number.[17]

Having received such a free brochure or referral, the individual is deemed to be a recipient of TANF services and “categorically eligible” for food stamps; the normal asset limits for food stamp eligibility are thereby waived.[18] Half of all food stamp recipients now enroll in the program through this procedure.[19]

In states using this loophole, a middle-class family with one earner who becomes unemployed for one or two months can receive $668 per month in food stamps even if the family has $20,000 in cash sitting in the bank. Because of this, food stamps has been transformed from a program for the truly needy to a routine bonus payment stacked on top of conventional unemployment benefits.

Categorical eligibility is an imprudent policy that automatically adds persons to the food stamp rolls without determining whether they are economically needy. An analysis conducted for the USDA by Mathematica Policy Research estimated that eliminating the asset limits in the food stamp program would expand program caseload and costs by 22 percent in good economic times.[20] In periods of very high unemployment, this number presumably would be quite a bit higher. Future food stamp expenditures could be substantially reduced if Congress eliminated categorical eligibility, restored normal asset limits on eligibility, and reestablished the gross income eligibility standard at 130 percent of the federal poverty level.

Another widely used loophole exploited by big-spending state governments is “Heat and Eat.” Food stamp benefit levels are based on “countable” income. The lower the countable income of an individual, the higher that person’s benefits will be. If a person or family pays for utility costs separately from rent, they can deduct both the utility cost and the rent from their countable income, thereby increasing the amount of food stamp benefits they receive.

The Heat and Eat loophole assigns a fictitious separate utility cost to food stamp recipients in order to increase their benefits. Under this loophole, current law allows states to average out their utility costs and use a Standard Utility Allowance (SUA) when determining food stamp eligibility. The law provides a higher SUA for individuals who pay for heat and air conditioning separately from their rent. However, a loophole in the law permits states to assign the higher SUA to anyone who receives aid from the Low Income Energy Assistance Program (LIHEAP) even if heating and air conditioning are already included in the individual’s rent or shelter deduction.[21] This allows the state welfare agency to falsely hike deductions for purposes of calculating countable income and thereby artificially boost the benefit levels.

Moreover, the law allows states to grant the higher SUA to food stamp applicants who receive only a tiny amount of LIHEAP aid. Consequently, states have begun issuing food stamps to applicants receiving LIHEAP benefits as low as $1 in order to boost their food stamp benefits. Because the federal government pays for the cost of both food stamps and LIHEAP, a state government loses no money from this sleight of hand and pulls in more food stamp spending. According to House estimates, closing this loophole would save taxpayers $14.3 billion over the next 10 years.[22]

4. Reduce fraud.

A recipient’s food stamp benefits decrease as monthly earnings increase. However, recipients can unlawfully receive excess benefits by failing to report new jobs or increased earnings to the food stamp office.

To reduce this type of fraud, state bureaucracies should be required, in a timely and consistent manner, to cross-check the National Directory of New Hires against their food stamp rolls to detect evidence of new employment. Recipients who have obtained employment but have failed to inform the food stamp office should have their benefits suspended. In addition, able-bodied adults receiving food stamps should be required to recertify for assistance every three months.

5. Prohibit food stamp payments to illegal immigrant families.

If an illegal immigrant gives birth to a child inside the United States, that child by law is an American citizen. As a result, the child automatically becomes eligible for food stamps and many other means-tested programs. There are roughly 4 million native-born children of illegal immigrants residing in the U.S.

Illegal immigrant parents routinely apply for welfare assistance for their children born in the U.S. The welfare office will clearly recognize that the immigrant is illegal and in violation of U.S. law but will take no action against the illegal immigrant. When the food stamp office provides aid for such a child, it gives the aid directly to the illegal immigrant parent in the form of an electronic benefit transfer (EBT) card. Although the benefits on the card are ostensibly limited to the U.S.-born child, these cards are used to purchase food for the illegal immigrant household, not just for the U.S.-born child.

This policy should be changed and brought in line with existing policy on the earned income tax credit. The EITC is a refundable, means-tested benefit for low-income families with children. In order to receive the EITC, parents must provide a valid Social Security number demonstrating that they are residing in the U.S. lawfully and are authorized to work. The same policy should be applied in food stamps: In the future, food stamps should be given only to children who reside with a parent, guardian, or foster parent who has demonstrated lawful U.S. residence.

6. Convert food stamps into a work activation program.

Food stamps is a fossilized program that, except for greatly increased costs, has changed little since its inception in the early years of the War on Poverty. For example, the program was largely unaffected by the welfare reform legislation of 1996, which replaced Aid to Families with Dependent Children with Temporary Assistance for Needy Families, even though TANF and food stamp caseloads overlap to a great degree.

Untouched by reform, it is an old-style entitlement program offering billions in unconditional aid. Recipients are entitled to one-way handouts and are rarely required to engage in constructive behavior as a condition for receiving that aid. Like the failed AFDC program, which it closely resembles, food stamps discourages work and rewards dependence.

There is a common misperception that the food stamp program is a program of temporary, short-term assistance. In reality, at any given moment, the majority of recipients are or will become long-term dependents. Historically, half of food stamp aid to families with children has gone to families that have received aid for 8.5 years or more.[23] (See Chart 2.)

Following the welfare reform model, food stamps should be transformed from an open-ended entitlement program that gives one-way handouts into a work activation program. Non-elderly, able-bodied adults who receive benefits should be required to work, prepare for work, or at least look for work as a condition of receiving aid.

Many food stamp households contain adults who are capable of working but work little or not at all. In the average month in 2010, 18.8 million households—roughly one household in five in the U.S.—received food stamp benefits.[24] Of this total, approximately 10.5 million households contained at least one able-bodied, non-elderly adult. This included around 7 million families with children and 3.5 million non-elderly, able-bodied adults without dependents (ABAWDs).

Among the 10.5 million food stamp households with able-bodied, non-elderly adults, 5.5 million performed zero work during the month.[25] Another 1.5 million to 2 million households had employment but appeared to work less than 30 hours per week. Altogether, each month, some 7 million to 7.5 million work-capable households received food stamps while performing no work or working less than 30 hours per week. These low levels of work are not simply the product of the current recession: They are typical of food stamp recipients even in good economic times.

A work activation program would seek to increase employment among able-bodied, nonworking food stamp households that do not work and to increase the hours of work among those who are employed part-time. Work activation should be phased in incrementally in the food stamp program when the current recession has ended. Typically, a work activation program will cause both the existing caseload and the number of new enrollments to drop rapidly.

A work activation program can operate at a fairly low cost. For example, a rigorous, closely supervised 16-week job search program would cost about $250 per recipient.[26] In one year, 10 million work-capable food stamp recipients could be circulated through this type of program at an annual cost of around $2.5 billion. This would cover all current work-eligible recipients who are nonworking or underemployed as well as many new work-capable enrollees.

Experience with welfare reform and the TANF program in the mid-1990s demonstrates that work activation can dramatically reduce welfare caseloads. In the four decades before welfare reform, TANF (formerly Aid to Families with Dependent Children) never experienced a significant decline in caseload. In the four years after welfare reform, the caseload dropped by nearly half.

Work activation could be designed to provide an incentive for states to reduce future dependence. If a state government operated its work activation program in a particularly effective way and reduced its food stamp caseloads below the pre-recession level, it might be allowed to retain a portion of the savings.

Use TANF funds to pay for implementation of a work requirement.

As noted, a work activation program will have administrative costs, but most states run their food stamp programs in tandem with their TANF programs, which already have a work requirement. Thus, the burden on states of implementing a work requirement for the food stamp population would not be as great as starting a separate work requirement from scratch.

Nonetheless, operating the work activation program will require additional funding. An appropriate funding source for a food stamp work activation program is the TANF program. Federal TANF funding is currently $16.5 billion per year, but only 40 percent of this funding is actually used to pay benefits. The other portion goes to a wide variety of other activities in state budgets.

Current TANF spending could be reduced by $2.0 billion per year, and these savings could be reallocated to fund a food stamp work activation program.[27] Reducing TANF spending to $14.5 billion would leave more than enough funding to cover the needs of the TANF population. The reallocated $2 billion would then be split among the states to cover the costs of instituting a new work activation requirement in the food stamp program.

7. Require drug testing as a condition of food stamp aid.

Means-tested welfare assistance should not be a one-way handout or an open-ended entitlement. Aid should be given on the basis of reciprocal obligation. Taxpayers should provide support to those in need, and recipients in return should engage in responsible and constructive behavior as a condition of receiving aid. Requiring welfare recipients to stop using illegal drugs is a core element of reciprocal obligation.

Welfare recipients are roughly twice as likely as the general public to use illegal drugs. Studies show that 21 percent of mothers receiving welfare reported using illegal drugs in the prior year.[28] Similarly, one-tenth of food stamp recipients report having used illegal drugs in the prior month, although the actual rate is probably far higher.[29] Self-reporting of illegal drug use almost certainly results in an undercount of actual use. The real rates of illegal drug use among welfare recipients may be five times higher than the self-reported rate.[30]

As welfare spending approaches $1 trillion per year, taxpayers have a right to know that their funds are being used frugally. Those who pay for the welfare state can properly insist that their aid go to those who are truly in need and not be wasted on frivolous or self-destructive activities, such as illegal drug use. In the case of food stamps, taxpayers should not be required to pay for food for individuals who waste their own money on illegal drugs.

Evidence shows that drug testing has the potential to significantly reduce unnecessary welfare spending and misuse of funds. For example, Florida’s policy of drug testing TANF applicants appears to have reduced new welfare enrollments by as much as 50 percent.[31] Potential applicants using illegal drugs simply chose not to enter the welfare system. Under the Florida policy, drug users could enroll in welfare in the future, but they would first need to stop using illegal drugs. The choice was theirs. During the time it was in operation, the Florida drug testing requirement produced $30.64 in reduced welfare costs for every dollar spent on its operation.[32]

The food stamp and TANF populations overlap to a considerable degree. In addition, the food stamp program includes ABAWDs, a population not in TANF. This group, which includes many single males, is likely to have high levels of drug use. Therefore, a drug testing requirement for food stamps would, like the Florida TANF reform, be likely to reduce caseloads significantly.

Finally, all welfare programs, including food stamps, should be designed to promote self-sufficiency among able-bodied adults and to discourage long-term dependence. Scientific evaluation of the Florida TANF drug testing requirement showed that earnings among welfare recipients who used illegal drugs were 30 percent lower than earnings among those who did not. Similarly, national data show that illegal drug use is twice as frequent among the unemployed compared to the fully employed.[33] (See Chart 3.)

Simply put, there is an overwhelming positive correlation between illegal drug use and lower levels of employment. Of course, it is possible that unemployment causes increased drug use: Unemployed persons could be depressed and therefore more likely to use drugs. In reality, however, causation is likely to run more heavily in the opposite direction, with illegal drug use leading to lower levels of employment. This would occur because users of illegal drugs often have attitudes, capacities, and habits that make them less likely to seek, obtain, and maintain high levels of employment.

Because illegal drug use is linked to lower levels of work, any serious effort to promote employment and self-sufficiency should include steps to discourage illegal drug use within the welfare population. A well-designed drug testing program would be an important tool in any effective welfare-to-work strategy.

Following the Model of Welfare Reform

These reforms roughly follow the model of the welfare reform of the mid-1990s. In 1996, Congress passed the Personal Responsibility and Work Opportunity Act (PRWORA), which replaced AFDC with TANF. One immediate result was plummeting caseloads. In the post–World War II period, the AFDC caseload had never experienced a significant decline, but after passage of welfare reform, the AFDC/TANF caseload dropped dramatically from 4.3 million families in 1996 to 2.2 million in 2000.[34] The caseload remains at a low level today.

As the caseload fell, employment of single mothers surged and child poverty dropped at an unprecedented rate. For example, before reform, the poverty rates of black children and children of single mothers had remained stubbornly frozen for 25 years, but after reform, both dropped quickly. The black child poverty rate fell from 41.5 percent in 1995 to 30 percent in 2001, and the poverty rate of children of single mothers dropped from 50.3 percent to 39.8 percent in the same years.[35]

The reform was successful in simultaneously reducing both dependence and poverty. The poverty rate for both groups has risen sharply during the current recession but remains below the pre-reform rates for recessionary periods.

The welfare reform of 1996 was based on three principles:

  1. The uncapped funding entitlement of the AFDC program was ended, and state governments were given a fixed sum of money in future years.
  2. The open-ended legal entitlement of recipients to cash payments based on fixed eligibility formulae was ended. Policymakers recognized that uncapped welfare entitlements tend to grow at a rapid and unsustainable pace.
  3. States were required to implement work activation programs for able-bodied TANF recipients.

These same principles should be used in reforming the food stamp program.

Block Grants Versus Work Activation. The TANF program, created by welfare reform, is often called a “block grant,” but that term is ambiguous and poorly understood. To block grant a program sometimes means (1) to eliminate an automatic legal entitlement to benefits for certain categories of persons and (2) to replace an uncapped entitlement spending mechanism with funding that is set at a fixed level and subject to budgetary controls.

However, “block grant” is also used in a second, much broader sense to mean a program in which fixed funds are collected at the federal level and turned over to state governments to spend with few or no requirements. TANF was never a block grant in this latter sense; indeed, one of its defining features was that, for the first time, it imposed significant work-based performance standards on the states. Rather than a block grant, TANF could more accurately be called a “work activation” grant.

Block grants in the second, broader sense are often viewed as vehicles of “federalism.” In reality, such block grants represent pseudo-federalism. Under pseudo-federalism, revenue is collected at the federal level and then turned over to state governments to be spent as the states choose. But collecting revenue at one level of government to be spent at another level of government is a recipe for inefficiency and non-accountability.

Pseudo-federalism is particularly inappropriate within the current means-tested welfare system in which the federal government provides more than three-quarters of the total funding. Real federalism, or turning welfare back to the states, would require states to pay for their own welfare programs with state revenues—something that no state is eager to do.

Didn’t Conservative President Ronald Reagan Champion Block Grants? It is true that during his first term in office, Ronald Reagan consolidated some 77 small separate government programs into nine block grants. Replacing myriad tiny categorical programs with larger, broader programs did give states greater flexibility. In some cases, uncapped entitlement spending was replaced by fixed funding, which slowed the growth of welfare spending somewhat.

But none of Reagan’s new block grant programs—which included the Community Development Block Grant, Community Services Block Grant, and Low Income Energy Assistance Program—was ever a source of policy innovation. Clearly, no one could ever accuse these programs of revolutionizing the welfare state.

In reality, Reagan’s vision in welfare went well beyond simple block grants. Reagan’s primary focus in welfare was to require able-bodied recipients to work, not to give states unlimited flexibility in spending federal revenue. It was Reagan’s emphasis on work that prompted the subsequent welfare reform revolution.


The U.S has spent $19.8 trillion on means-tested welfare since President Lyndon B. Johnson launched the War on Poverty in the 1960s. Spending on food stamps alone has totaled $1.2 trillion.[36] As noted, in 2011, government spent $927 billion on means-tested assistance. This amounts to nearly $9,000 per year for each poor and low-income American.

In the short term, much of this spending props up the living standards of the poor. Not even the government can spend $9,000 per person without significantly affecting living conditions. But the original goal of the War on Poverty was not to prop up living standards artificially through an ever-expanding welfare state. When President Johnson launched the War on Poverty, he declared that it would strike “at the causes, not just the consequences of poverty.”[37] He added, “Our aim is not only to relieve the symptom of poverty, but to cure it and, above all, to prevent it.”[38]

In other words, President Johnson was not proposing a massive system of endlessly increasing welfare benefits doled out to an ever-enlarging population of beneficiaries. His proclaimed goal was not to create a massive new system of government handouts, but an increase in self-sufficiency: to create a new generation of Americans capable of supporting themselves out of poverty without government handouts. LBJ planned to reduce, not increase, welfare dependence. The goal of the War on Poverty, he stated, would be “making taxpayers out of taxeaters.”[39] He declared, “We want to give the forgotten fifth of our people opportunity not doles.”[40]

However, in terms of reducing the causes rather than the consequences of poverty, the War on Poverty has failed utterly. After $19.8 trillion in spending, the situation is worse, not better. A significant portion of the population is now less capable of prosperous self-sufficiency than when the War on Poverty began.

Now President Obama has called for another permanent increase in the welfare state. He plans to spend $12.7 trillion on means-tested aid over the next decade—roughly $270,000 for each current poor person in the nation. Much of this spending would be funded by borrowing from abroad and putting future generations further in debt.

Obama’s plans for a permanent expansion of the welfare state are unsustainable. Future spending needs to be subject to reasonable limits. When the economy recovers, total means-tested spending should be returned to pre-recession levels adjusted for inflation. Food stamp spending should also be returned to pre-recession levels when the recession ends. In addition, when the economy improves, able-bodied, non-elderly adults receiving food stamps should be required to work, prepare for work, or at least look for a job as a condition of receiving aid.

Robert Rector is Senior Research Fellow in the Domestic Policy Studies Department, and Katherine Bradley is a Visiting Fellow in the Richard and Helen DeVos Center for Religion and Civil Society, at The Heritage Foundation.


Categorical Eligibility


The food stamp program gives state governments broad flexibility to determine benefit eligibility within the state by selecting different standards of categorical eligibility. According to USDA data, states may select one of three different categorical eligibility options. [41]

  1. Traditional categorical eligibility. Under traditional categorical eligibility, a household where all members are eligible to receive means-tested cash aid from Supplemental Security Income, Temporary Assistance to Needy Families, or General Assistance automatically becomes eligible for food stamp benefits as well. This gives states some flexibility because they can set the standards for receipt of cash benefits from these three programs; however, households that do not receive cash aid from these programs must apply for food stamp aid through the normal food stamp application process and are therefore subject to the normal food stamp limits on assets and gross income.
  2. “Narrow” categorical eligibility. Under this option, a household will become categorically eligible for food stamps if it receives cash aid from the three programs listed above or if a member receives an actual non-cash social service funded by TANF such as day care assistance, transportation aid, or general counseling.
  3. Broad-based categorical eligibility. Under this option, a household becomes categorically eligible for food stamp benefits whenever it receives a TANF-funded brochure or pamphlet from the state food stamp office.

Since the federal government pays for 100 percent of the cost of food stamp benefits, state governments have a financial incentive to enroll as many individuals as possible in the program. As of January 2012, in operating their food stamp programs, five states used traditional categorical eligibility standards; five states used the narrow categorical eligibility standard; and 40 states and the District of Columbia used the broad-based categorical eligibility option.

See Appendix Table 1 for a description of the options employed by specific states.



[1]Federal outyear spending was taken from U.S. Office of Management and Budget, Budget of the United States Government, Fiscal Year 2013: Analytical Perspectives (Washington: U.S. Government Printing Office, 2012), Table 32-1, (accessed June 21, 2012). The figure in the text includes estimated state administrative expenditures equaling 9 percent of total federal spending.

[2]This figure includes both projected federal expenditures and projected state contributions to federal means-tested programs. Federal expenditures were taken from ibid. For additional information, see Robert Rector, “Examining the Means-Tested Welfare State; 9 Programs and $927 Billion in Annual Spending,” testimony before the Committee on the Budget, U.S. House of Representatives, April 17, 2012,

[3] Ibid.


[5]This calculation combines potential welfare aid with non-welfare income currently received by the poor.

[6]To avoid abrupt changes, spending could be reduced incrementally over a two-year period.

[7]The food stamp program is technically a capped entitlement program rather than a pure entitlement program; however, the level of maximum spending is set high enough that it does not limit outlays. Consequently, the food stamp program operates like an ordinary uncapped entitlement program.

[8]Peter H. Rossi, Feeding the Poor: Assessing Federal Food Aid (Washington: American Enterprise Institute Press, 1998), p. 102.

[9]Author’s calculation derived as follows: Each $1.00 in food stamp benefits received by a household increases food consumption expenditures in the household by around 30 cents. Ibid., p. 4. According to the U.S. Department of Commerce National Income and Product Accounts, farmers’ income represents about one-tenth of the cost of food and beverages consumed in the home in the U.S. Consequently, we can assume that each dollar in food stamp spending will result in around three cents in added farmers’ income.

[10]Gene Falk and Randy Alison Aussenberg, “The Supplemental Nutrition Assistance Program: Categorical Eligibility,” Congressional Research Service Report for Congress, March 2, 2012, (accessed June 21, 2012).

[11]The Bush Administration repeatedly but unsuccessfully sought legislation to overturn the Clinton regulations on broad-based categorical eligibility. Ibid. p. 1.

[12]The TANF program is a program with fixed funding that replaced the older Aid to Families with Dependent Children (AFDC) program, which was funded as an open-ended entitlement. Under AFDC, states were given more funds if they increased the number of persons enrolled in the program. Because of this, the federal government maintained tighter control of program eligibility. By contrast, in the TANF program, states are given a fixed annual funding that is not increased if the state increases enrollments. Under this fixed funding principle, states were given very wide flexibility in determining who would receive TANF cash aid and services.

[13]To become eligible to receive food stamp aid, individuals or households are subject to three criteria: an asset limit, a gross income limit, and a net or countable income limit. The categorical eligibility loophole waives the first two limits, but the value of benefits received will still be determined by the level of countable income.

[14]Retirement assets such as IRAs and educational savings funds are excluded from the count of liquid assets.

[15]Joe Richardson, “The Federal Response to Calls for Increased Aid from USDA’s Food Assistance Program,” Congressional Research Service Report for Congress, February 17, 2010, (accessed June 21, 2012).

[16]Falk and Aussenberg, “The Supplemental Nutrition Assistance Program: Categorical Eligibility.”

[17]Ibid. p. 6. MOE stands for state maintenance of effort funds. These are state funds that the state is required to contribute to TANF or a broad variety of TANF-related activities.


[19]According to the Congressional Budget Office, “In fiscal year 2010, about 50 percent of SNAP participants were considered eligible for benefits through their receipt of noncash TANF benefits, so they were not subject to the same income and asset requirements as other participants.” Congressional Budget Office, “The Supplemental Nutrition Assistance Program,” April 2012, p. 8, (accessed July 11, 2012).

[20]Karen Cunnyngham and James Ohls, Simulated Effects of Changes to State and Federal Asset Eligibility Policies for the Food Stamp Program, U.S. Department of Agriculture, Food and Nutrition Service, Contractor and Cooperator Report No. 49, October 2008, p. xvi, (accessed July 11, 2012).

[21]LIHEAP is the federal government’s welfare program that pays for low-income families’ heating and energy costs.

[22]Report 112-470, Sequester Replacement Reconciliation Act of 2012, Report of the Committee on the Budget, U.S. House of Representatives, to Accompany H.R. 5652, 112th Cong., 2nd Sess., May 9, 2012, p. 20, (July 11, 2012).

[23]Robert Rector, “Reforming Food Stamps to Promote Work and Reduce Poverty and Dependence,” testimony before the Committee on Agriculture, U.S. House of Representatives, June 27, 2001, (accessed June 22, 2012).

[24]U.S. Department of Agriculture, Food and Nutrition Service, Characteristics of Supplemental Nutrition Assistance Program Households: FY 2010, September 2011, (accessed June 21, 2012).

[25]The figure of 5.5 million was derived as follows: Approximately 4.6 million families with children had no earnings. Of these, probably 1.8 million did not contain an able-bodied adult, leaving some 2.7 million families with able-bodied adults but no work. In addition, 2.8 million (80 percent) of the 3.5 ABAWD households had no earnings in the month. See ibid., p. 52, Table A16.

[26]Details of this estimate are available from the authors upon request.

[27]Additional funds could be reallocated from the existing food stamp training program.

[28]Rukmalie Jayakody, Sheldon Danziger, and Harold Pollack, “Welfare Reform, Substance Abuse and Mental Health,” Journal of Health Politics, Policy and Law, Vol. 25, No. 4 (August 2000), pp. 623–652, (accessed June 21, 2012).

[29]U.S. Department of Health and Human Services, Substance Abuse and Mental Health Services Administration, Office of Applied Studies, “Substance Use Among Persons in Families Receiving Government Assistance,” The NHSDA Report, April 19, 2002, Table 1, (accessed June 21, 2012).

[30]Jayakody et al., “Welfare Reform, Substance Abuse and Mental Health,” p. 11.

[31]Tarren Bragdon, “Florida’s Drug Test Law for Welfare Cash Assistance: First Quarter Facts,” Foundation for Government Accountability, October 13, 2011, (accessed June 21, 2012).

[32]Ibid. Judge Scriven, who struck down the Florida drug testing law, refused to consider the strong evidence that the program substantially reduced new welfare enrollments.

[33]U.S. Department of Health and Human Services, Substance Abuse and Mental Health Services Administration, Results from the 2010 National Survey on Drug Use and Mental Health: Summary of National Findings, September 2011, p. 23, (accessed June 21, 2012).

[34]Robert Rector and Patrick F. Fagan, “The Continuing Good News About Welfare Reform,” Heritage Foundation Backgrounder No. 1620, February 6, 2003,


[36]Both figures are in constant 2011 dollars.

[37]Lyndon B. Johnson, “Proposal for a Nationwide War on the Sources of Poverty,” March 16, 1964, (August 27, 2009).

[38]Lyndon B. Johnson, “Annual Message to the Congress on the State of the Union,” January 8, 1964, (August 27, 2009).

[39]Quoted in David Zarefsky, President Johnson’s War on Poverty: Rhetoric and History (Tuscaloosa: University of Alabama Press, 1986), p. 49.


[41]The information in this appendix is drawn from Falk and Aussenberg, “The Supplemental Nutrition Assistance Program: Categorical Eligibility,” and U.S. Department of Agriculture, Food and Nutrition Service, Supplemental Nutrition Assistance Program: State Options Report, 9th ed., November 2010, (accessed July 1, 2012).


Katherine Bradley
Katherine Bradley

Former Research Fellow, Domestic Policy Studies

Robert Rector
Robert Rector

Senior Research Fellow, Center for Health and Welfare Policy