SECTION: CAPITOL HILL HEARING TESTIMONY

LENGTH: 7872 words

COMMITTEE: SENATE AGRICULTURE, NUTRITION AND FORESTRY

HEADLINE: 2002 FARM BILL

TESTIMONY-BY: MR. ROBERT GREENSTEIN, EXECUTIVE DIRECTOR,

AFFILIATION: CENTER ON BUDGET AND POLICY PRIORITIES

BODY:
July 19, 2001

TESTIMONY OF

MR. ROBERT GREENSTEIN Executive Director, Center on Budget and Policy Priorities

before the Senate Committee on Agriculture, Nutrition, and Forestry

I would like to thank the Chairman and the Members of the Committee for the invitation to testify. I am Robert Greenstein, director of the Center on Budget and Policy Priorities, an organization that conducts research and analysis both on fiscal policy and on policies and programs focused on low- and moderate- income families. Many years ago, I had the opportunity to serve as Administrator of the Food and Nutrition Service at USDA. My testimony today focuses on the food stamp program, a program in which I've had a keen interest for nearly 30 years. Background on the Food Stamp Program

The food stamp program has long been unique among low-income programs. It is the only major benefit program that covers all types of low-income households. It has always differed from welfare programs; working-poor families and two-parent families have always been eligible, rather than being ineligible or subject to more restrictive eligibility rules. In addition, the food stamp program is the sole major low-income program that places poor children and their families in the same eligibility- and-benefit structure as the low-income elderly and disabled, a feature of the program that has been beneficial to poor families with children, since they usually are accorded less favorable treatment than the elderly and disabled when separate programs or separate eligibility-and-benefit rules are established.

The food stamp program also is the nation's most important child nutrition program -- half of all food stamp recipients are children, and the program provides more food assistance to low- income children than the school lunch program or any other child nutrition program. A team of doctors the Field Foundation sponsored that examined hunger and malnutrition among poor children in the South, Appalachia, and other poor areas in 1967 (before the food stamp program was widespread in these areas) and again in the late 1970s (after the program had been instituted nationwide) found marked reductions over this period in hunger and malnutrition among children. The doctors attributed these improvements in large part to the food stamp program; when the doctors reported their findings to this Committee in the late 1970s, they stated: "The food stamp program does more to lengthen and strengthen the lives of disadvantaged Americans than any other non-categorical social program" and "is the most valuable health dollar spent by the federal government." Findings such as these led Senator Robert Dole, during his tenure in the Senate, to term the food stamp program the most important social program since Social Security.

Another significant feature of the program is that it targets benefits to need, seeking to provide benefits commensurate with a household's ability to purchase a basic diet. Households with the lowest disposable incomes -- measured as income after certain expenses -- receive the largest benefits. As income rises, benefits are reduced. In addition to improving the efficiency of the program, this program feature serves another purpose -- it reduces inequities between households that secure certain other government benefits and those that do not. A working-poor family with high child care or housing costs receives more food stamps than a family that has similar earnings but does not incur such costs because it receives a publicly funded child care or housing subsidy.

Recent Changes in the Composition of the Food Stamp Caseload

Over the past decade, there have been rather dramatic changes in who receives food stamps. Stated simply, it has become much less a program for non-working families receiving public assistance and much more a program for working households.

Consider the following figures. In FY 1989, nearly 60 percent of all families with children receiving food stamps were families that received AFDC and had no earnings. In FY 1999, only 37 percent of food stamp families with children had these characteristics. The figure is certain to be still lower today.

Similarly, 10 years ago, the number of food stamp households with children that received AFDC and had no earnings was more than double the number of working food stamp households with children. Today, the number who work exceeds the number who receive cash welfare benefits and are not employed.

Indeed, by fiscal year 1999, only 13 percent of food stamp recipients were adults who were not elderly, disabled, or caring for an incapacitated person and who lived in a household in which no one was employed. (1) The large majority of this 13 percent of recipients -- about two-thirds of them -- consisted of individuals participating in TANF and hence subject to TANF work requirements or participating in another employment or training program. Only four percent of all food stamp participants consisted of adults who were not elderly or disabled, caring for an incapacitated person, subject to TANF work requirements, or participating in another employment and training program.

Moreover, about half of this remaining four percent of participants were mothers caring for a young child. Five-sixths of this four percent of participants were mothers caring for a young child, individuals registered for work in the food stamp program, individuals exempt from work registration for other reasons, or individuals without children whose tenure on food stamps is limited to three months out of each three-year period. The majority of the very small number of remaining recipients were recently unemployed workers who were receiving unemployment insurance or individuals enrolled in an educational program at least half time. Only two-tenths of one percent of food stamp recipients did not fall into one of these categories.

Trends in Food Stamp Participation

Food stamp participation has fallen dramatically in recent years. In 1994, some 27.5 million people received food stamps in an average month. In 2000, 17.2 million did. This is the sharpest decline in the program's history.

The robust economy clearly was a major factor here. But more than the economy was involved. As Figure 1 indicates, from 1994 (when food stamp participation peaked) to 1999 (the latest year for which poverty data are available), the number of people living in poverty declined 16 percent, (2) while the number of people receiving food stamps fell 35 percent. The figures for children are similar -- the number who are poor fell 21 percent (an impressive accomplishment) but the number receiving food stamps fell 36 percent. The loss of food stamps among many families that remained poor is a key reason why the Census data show that over this period, the children who remained poor became poorer, on average. In 1999, the average poor child fell farther below the poverty line than in any year since the Census Bureau began collecting those data in 1979. (3)

Part of the decline in food stamp participation that was not attributable to the economy resulted from provisions of the 1996 welfare law that eliminated food stamp eligibility for most legal immigrants and limited food stamp eligibility to three months out of each three-year period for non-disabled adults aged 18 to 50 who are not raising minor children and are unemployed. This eligibility changes can, however, explain only a fraction of the large food stamp participation decline.

The evidence is now clear one of the two largest factors behind the food stamp participation decline (the economy being the other) was a large, unexpected drop in the proportion of low- income households eligible for food stamps that actually receive them. Studies conducted for USDA by Mathematica Policy Research show that in 1994, some 71 percent of those eligible for food stamps secured them, but in 1998, only 59 percent did. (1998 is the latest year for which these data are available.) The decline was particularly large among eligible households with incomes between 50 percent and 100 percent of the poverty line, a category where most of the working poor are found.

The decline in participation rates has been especially large among children. In 1994, 86 percent of eligible children received food stamps. In 1998, some 69 percent did.

Reductions in Food Stamp Costs

Food stamp costs have fallen sharply in recent years. Food stamp benefit expenditures in fiscal year 2000 were $15 billion. This represented a 23 percent decline from expenditures in fiscal year 1994, after adjustment for inflation.

Some of this reduction resulted from provisions of the welfare law. When the law was enacted, CBO estimated its food stamp provisions would reduce food stamp expenditures by nearly $28 billion over six years and account for half of the savings in that law. CBO also estimated these provisions would eliminate nearly one-fifth of projected food stamp benefit expenditures by 2002.

The CBO estimates showed that a substantial majority of these savings stemmed not from the law's provisions restricting eligibility for legal immigrants and adults without children, or from the provisions of the law that toughened food stamp requirements related to work, payment of child support, and the like. Instead, the CBO estimates show that two-thirds of the food stamp savings resulted from provisions to reduce food stamps benefits, often on an across-the-broad basis, for most or all categories of food stamp households, including the working poor and the elderly and disabled. Although a few of these benefit- reduction provisions subsequently were scaled back, most remain in effect. A typical family of four with a parent who works full- time year-round at the minimum wage now receives $240 a year less in food stamps as a result of these changes.

In other words, for most poor households, the food purchasing power of food stamp benefits has eroded. Further erosion will occur in coming years because of a provision of the welfare law that eliminated an important inflation adjustment in the program. (This Committee originally rejected that provision; it was eventually included in conference.) Over time, this provision will lessen the program's effectiveness in helping low-income families purchase a nutritionally adequate diet.

These provisions were included in the welfare law to meet austere budget targets the Agriculture Committees were assigned during a period when large deficits were believed to loom indefinitely. These across-the-board or near across-the-board benefit cuts had no relation to welfare reform goals such as promoting work and marriage. As a result of these provisions, budget cuts were deeper in the latter half of the 1990s in the food stamp program than in any other major social program in the federal government.

The combined effect of the economy, the food stamp budget cuts, and the unexpected decline in food stamp participation rates has produced the large reduction in food stamp expenditures noted above. Figure ____ helps to illustrate this point. The top line in the figure shows CBO's estimate, prior to the welfare law's enactment, of what food stamp costs would be in fiscal years 1997 through 2000 if the Food Stamp Act remained unchanged. The middle line shows CBO's estimate, at the time of the welfare law's enactment, of food stamp expenditures in 1997-2000 under the welfare law; this line includes downward adjustments we have made in the CBO cost estimate to reflect the fact that the economy has performed better since 1996 than CBO had forecast. The bottom line in the chart shows actual food stamp expenditures in these years. The large and growing gap between the middle line and the bottom line primarily reflects the large and unexpected declines in food stamp participation rates.

What Has Caused the Unexpected Decline in Participation Rates?

To assess possible reforms in the food stamp program, two additional issues need to be examined: what has caused the large decline in food stamp participation rates?; and how is the current operation of the food stamp program affecting the welfare- reform goal of promoting work?

With regard to the trends in participation rates, it should be noted that eligible working families have always had a lower food stamp participation rate than families receiving public assistance. Hence, the combination of increases in employment and decreases in welfare receipt would be expected to result in some decline in the overall food stamp participation rate. In addition -- and of particular concern -- the participation rate among working families with children has itself declined. Between 1994 and 1998, the percentage of eligible working households with children that receive food stamps fell from 59 percent to 51 percent. The program began doing less well at serving working poor families at the very time that welfare reforms sought to encourage poor families to go to work rather than rely upon cash welfare assistance.

Virtually all eligible households receiving TANF cash benefits receive food stamps. By contrast, only half of the eligible working-poor households do. The movement from welfare to work has heightened the importance of this low participation rate among eligible working-poor households. Policymakers of both parties have for some time espoused the principle that if a parent works full time throughout the year, his of her children should not have to be in poverty. With expansions in the Earned Income Tax Credit under Presidents Reagan, Bush, and Clinton, this goal has nearly been achieved for a family of four or fewer people -- the combination of full-time minimum wage earnings, the EITC, and food stamps lift a family of four just about to the poverty line. If, however, such a family does not receive food stamps despite being eligible for them, it still falls a substantial distance below the poverty line. (See Figure ___.)

Urban Institute studies shed some light on one aspect of this matter. These studies show that large numbers of families working their way off welfare but remaining poor are losing food stamps when they go to work, despite remaining eligible for them. The Urban Institute examined families with children that received welfare and food stamps in 1997 but had left the welfare rolls by 1999. The researchers found that in 1999, only 43 percent of the "welfare-leaver" families that had incomes below the food stamp income limits were receiving food stamps. Urban Institute researchers also found that one-third of the families that left welfare during this period reported they had either cut the size of meals or skipped meals in the past year because they did not have enough food.

Various studies have documented problems in the interaction between TANF and food stamps when a family leaves welfare for work. Large numbers of families apparently do not realize their food stamp eligibility continues at this point. Lack of information about the differences between food stamp and TANF eligibility rules is, however, only one of the factors related to the drop in food stamp participation rates. Other factors also are playing a role.

Of particular importance are changes in many areas in the food stamp administrative procedures that low-income working families must navigate to secure and retain food stamp benefits. The Urban Institute studies found that between 1997 and 1999, the number of eligible, non-participating families that cited administrative problems as a reason for their non-participation nearly doubled.

For example, in recent years, a number of states have shortened the duration of time for which working families (and in some cases, other families as well) may be certified for food stamps and required such families to return to food stamp offices at more-frequent intervals to reapply. This can cause parents to miss time from work and forgo some wages. It also can aggravate employers. Between 1994 and 1999, a dozen states increased by 50 percentage points or more the proportion of working households with children that were assigned food stamp certification periods of three months or less (i.e., required to reapply at three-month intervals to retain their benefits). In these states, the number of working households with children that receive food stamps fell 29 percent over this five-year period. In the other states, the number of working households with children receiving food stamps declined less than one percent. Similarly, overall food stamp participation rates fell twice as much in states that greatly increased the use of these short certification periods for working households as in other states.

The spread of such practices has stemmed not from the welfare law or the pursuit of welfare-reform goals but from intensified state efforts to lower food stamp error rates, induced by increased federal pressures on states to reduce errors. Under the food stamp quality control system, any state with an error rate above the national average is potentially subject to fiscal sanctions. State performance on error rates is the sole performance measure the food stamp law sets. States often measure the performance of local food stamp offices, supervisors, and even individual eligibility workers based partly on the error rate of the cases they handle.

The substantial increase in recent years in the proportion of the food stamp caseload that consists of households with earnings has greatly intensified these pressures on states. Because household earnings can fluctuate and food stamp benefits are sensitive to modest changes in monthly income or expenses (each change of $3 in net household income results in roughly a $1 change in benefits), precise benefit accuracy is more difficult to achieve for households with earnings than for households receiving public assistance. Welfare payments usually remain constant from month to month, and if a family's welfare grant level does change, the welfare office -- which administers both welfare and food stamps -- knows of the change immediately. Among working poor households, by contrast, hours of work and hence household earnings can fluctuate due to changes in employers' needs, temporary breakdowns in child care arrangements, or a parent or child becoming sick. (Many low-wage jobs do not provide paid sick leave.) It is difficult for food stamp offices (and households) to track these earnings fluctuations on a month-to-month basis. Until recently, if a working poor parent worked just one hour more or less per week than the food stamp office had projected, the state was charged with an error. The average error rate in 1998 was nearly twice as high among households with earnings as among those without earnings.

The recent increases in the share of the caseload consisting of working families thus have exerted upward pressure on state error rates. Indeed, the more effective a state's welfare reform program is in moving families from welfare to work, the greater the risk the state has of seeing its food stamp error rate increase and of becoming subject to federal financial sanctions. (4)

Many states responded by instituting aggressive procedures to reduce the potential for errors among working households, including shorter certification periods and more intensive paperwork and verification requirements. In addition, because of quality control pressures, some states subject categories of households considered to be "error prone" to more exhaustive procedures; in some areas, simply having earnings can place a household in an error-prone category. These practices make it more difficult for families to secure and retain food stamps if they go to work.

Citizen Children in Immigrant Families

Another factor that has contributed to decreases in participation rates has been a remarkable decline in food stamp receipt among children who are U.S. citizens but live in a household that contains one or more persons (usually a parent) who is a legal permanent resident. USDA data show that between 1994 and 1998, the number of such citizen children receiving food stamps fell 74 percent, despite the fact that children who are citizens did not lose food stamp eligibility. The number of these citizen children who receive food stamps plunged by one million, from 1.35 million children in 1994 to 350,000 in 1998. The participation rate among citizen children living in a household with a non-citizen member fell from 76 percent to 38 percent during this period.

This decline appears attributable in part to confusion about the complex food stamp rules that apply to immigrant families. Legal immigrant parents are ineligible for food stamps even if they entered the country before the welfare law was signed, while their children are eligible if they either are citizens or entered the country before August 22, 1996. Apparently, some parents mistakenly believe their children are ineligible along with themselves.

How is the Food Stamp Program Affecting State Efforts to Promote Movement from Welfare to Work?

At various points over the past several years, two very different criticisms have been voiced regarding the effect of the food stamp program on welfare-reform efforts to promote work. The criticisms were based largely on anecdotal information and personal belief and intuition, since not much data or research had been completed in this area. Now more research and data are available to help evaluate these issues.

The first criticism is that the food stamp program may be compromising welfare-reform efforts to promote work because the presence of the food stamp program may be inducing significant numbers of families with children to escape TANF work requirements by leaving or failing to enroll in TANF and enrolling only in food stamps. If such an effect were present, there would be substantial increases since the enactment of TANF in the number of food stamp households with children that are neither working nor receiving welfare. The rich data that USDA makes available on the food stamp program can be examined to determine if this has been occurring.

The good news is that the data show this has not occurred. Between 1996, when the welfare law was enacted, and 1999 (the latest year for which these data are available), the number of food stamp families with children that neither work nor receive cash welfare benefits declined rather than increased, falling by more than 90,000. The decline occurred among both two-parent and single-parent families.

These figures are significant. Welfare rolls have plummeted in recent years, and studies have consistently shown that a significant fraction of those who have ceased receiving welfare have done so without becoming employed. As a result, one would have expected to see some increase in the number of families that receive food stamps but are neither employed nor on welfare. The fact that this did not occur and that the number of such families has declined strongly indicates that the food stamp program is not hindering TANF work requirements and that families are not seeking to avoid working by foregoing cash welfare benefits and relying only on food stamps.

That this did not occur may be related, in part, to several changes made in the food stamp program in 1996. Before then, a family's food stamp benefits increased if it lost income because its welfare benefits were sanctioned due to noncompliance with a work requirement. As a result of the food stamp policy changes made in 1996, food stamp benefits no longer increase when welfare benefits decline because of a sanction. In addition, the welfare law stiffened the food stamp program's own sanctions for noncompliance with work and other behavioral requirements, including noncompliance with a TANF work requirement. Now, instead of a household's food stamps increasing if a family is sanctioned in TANF, its food stamps may be reduced.

A second likely reason that the number of food stamp families with neither employment nor welfare benefit has not increased -- and that non-working families do not appear to be bypassing TANF to avoid its work requirements and enrolling only in food stamps -- is that food stamps do not enable families lacking both earnings and welfare benefits to pay rent or utility bills or meet any other necessities that require cash.

The second criticism related to the food stamp program's possible effects on welfare-reform efforts is a very different one. This criticism is that by not serving the working poor adequately, the food stamp program is lessening the incentive to move from welfare to work. If families on welfare believe they will lose their food stamp benefits as well as their welfare check if they go to work at low wages, the net gain from working may be too small to make work attractive.

There is some evidence to support this criticism. Studies by both Mathematica Policy Research and the Manpower Demonstration Research Corporation find that many welfare families do not understand they will remain eligible for food stamps if they leave welfare for low-wage work. A recent MDRC paper specifically warns that this may be lessening incentives to move from welfare to work and thereby weakening welfare-reform efforts. Many state officials have urged that steps be taken to improve the food stamp program's performance in serving low-income working families not on welfare.

Food Stamp Reforms

Improvements can be made in the food stamp program to help address these issues. Changes are needed to retool the program to improve access by the working poor -- including program simplification, reform of the quality control system, and the granting to states of more flexibility over various aspects of the delivery of benefits to eligible households. Changes also are needed to address the overly large reductions of recent years in the food purchasing power the program provides to the working poor, the elderly, and other households, and to narrow gaps in coverage.

A piece of bipartisan legislation recently has been introduced that addresses some of these issues -- S. 583, introduced by Senators Kennedy, Specter, Leahy and other senators (and its House companion bill, H.R. 2142, introduced by Rep. James Walsh, Eva Clayton, and a number of other House members from both parties). This legislation would restore benefits to legal immigrants, improve the adequacy of food stamp benefits, ease the transition from welfare to work, and establish several pilot projects to test various other possible program reforms. The provisions of this legislation are sound and warrant the Agriculture Committee's careful consideration.

S. 583 does not address all of the matters needing attention in reauthorization, however. Other measures are needed in such areas as program simplification and quality control reform. The following discussion examines an array of provisions not included in S. 583 that merit consideration.

I. Improving Access by the Working Poor

A. Reform of the System for Measuring State Performance

Governors, state administrators, and independent analysts have urged an overhaul of the food stamp quality control system. The current system penalizes states for serving working families. Perversely, the larger the proportion of a state's caseload that consists of welfare families and the smaller the proportion that consists of working families, the lower a state's measured error rate will be. As a result, the current system can drive states to institute procedures that impede participation in the food stamp program by working households, compel such households to have to take time off from work at frequent intervals to maintain their food stamps, and essentially treat working families less favorably than other families. Adding to this problem, the food stamp quality control system makes states subject to sanction if their error rate is above the national average, which means that about half of the states are subject to sanctions each year, even if states as a whole are performing well and have achieved a low national error rate. Indeed, that is what has been occurring.

The food stamp quality control system was largely constructed a number of years ago when food stamp error rates were much higher than they are today. When food stamp quality control sanctions first were enacted, the food stamp overpayment rate was 17 percent. Now, with the advent of computers, better trained staff, and other administrative improvements, the overpayment rate has fallen to 6.5 percent. Moreover, a large share of the errors that do occur are errors that result when a low-income household that is eligible for food stamps receives a modestly larger or smaller benefit than it should. Only two percent of households receiving food stamps should have been found ineligible for the program. In short, while there is room for further improvement, primarily in a handful of states with high error rates, achievement of a national overpayment rate of 6.5 percent in a program as large and complicated as this one reflects strong performance. The severity of the current quality control system, which operates as a blunt instrument on the states, is not appropriate in the current environment.

Cash assistance and Medicaid used to have quality control systems similar to that which the food stamp program still employs. They no longer do. With 100 percent federal funding of food stamp benefits, some form of food stamp quality control system is necessary, but there is widespread consensus that substantial changes are needed here.

In the late 1980s, a blue-ribbon National Academy of Sciences panel studied the food stamp quality control system at USDA's request. The panel's report provides a sound starting point for thinking about QC system reform. One of the panel's principal recommendations was to cease basing sanctions on measures that subject large numbers of states to sanctions and instead to adopt a standard that focuses sanctions on states whose error rates make them "outliers." The worst state performers should be subject to sanctions, the panel recommended, rather than close to half of the states.

In addition, a more comprehensive assessment of states' performance in operating the program is needed. Payment accuracy should continue to be at the core of the system for measuring state performance, but payment accuracy should not be the sole measure the federal government uses in measuring performance. Other measures, such as measures related to performance in serving eligible households, and especially working families, also should be employed, with possible fiscal incentives attached.

Such improvements in the quality control and performance measurement system could have strong beneficial effects. In the past two years, USDA has increased state flexibility in certain areas to enable states to remove various barriers to participation by the working poor. Some states have been wary of taking advantage of this flexibility, however, since the QC system can effectively penalize them for improving service to the working poor.

Program developments in recent years indicate that when state flexibility is coupled with rigid and unrealistic quality control rules, states can feel compelled to institute measures that have the effect of making food stamps more difficult for eligible working families to secure. By contrast, if greater state flexibility over the delivery of benefits is coupled with significant quality control reforms, states are more likely to institute measures that improve service to eligible low-income working families. That has been the experience in Medicaid in recent years -- an easing of quality control pressures and added state flexibility over the delivery of benefits have resulted in changes in states that have made that program more accessible to the working poor.

B. Simplification

Various rules related to food stamp eligibility and benefit determinations are too complex for a program aimed at working families, especially certain rules related to small amounts of income. To ensure that no family receives a dollar more than it is considered to need, federal rules get into minutia such as when small amounts of money received from giving blood count as income. Some of these intricate rules are in law; others are in regulations. These rules make little difference in the total amount of benefits provided, but they complicate the program for states and households alike, and thereby increase both errors and barriers to participation. Such rules should be reviewed and cleaned out, with an eye to reducing the number of minor forms of income and assets that must be assessed.

Some simplification of the program's deduction structure also may be possible. Efforts in this area should be sensitive to the impact that such changes would have on those who benefit from the deductions. The deductions are designed to enable states to determine the amounts that households have available to buy food, so that benefits can be targeted on need. Modifying the deductions with the sole goal of simplification could have deleterious impacts on millions of individuals, causing some of the neediest households to face substantial benefit cuts. That having been said, some simplification is both possible and desirable.

For example, if a Medicare prescription drug benefit is enacted and the component of Medicaid that pays for Medicare premiums, deductibles, and co-payments for poor elderly and disabled is improved so it reaches more of those eligible, elimination of the food stamp excess medical expense deduction would be a possibility. (Without such an improvement in prescription drug coverage, eliminating the medical deduction would result in reducing benefits for nearly 300,000 elderly and disabled participants.)

In addition, some have raised concerns that the food stamp shelter deduction is unnecessarily complex. Measures should be explored to simplify this deduction without eliminating it or compromising its role in targeting benefits on families whose housing costs significantly reduce the income they have available for food. While the use of an average deduction across households would cause serious hardship and not be desirable, the deduction can be made simpler for households and less error-prone for states.

A significant part of the deduction's complexity stems from the rules for calculating households' utility costs. Although rent costs tend to remain unchanged from month to month, utility costs fluctuate. Current law seeks to simplify the determinations of utility costs and to avoid monthly benefit adjustments by allowing states to use statewide estimates called standard utility allowances (SUAs) in lieu of determining each household's actual utility costs. Unfortunately, the law imposes limitations on when the SUA may be applied that, while analytically rational, greatly undermine the SUA's capacity to simplify the calculation of the deduction. Although these rules do not affect terribly large numbers of households, they markedly increase the complexity of the procedures that states must teach their eligibility workers and the instructions they must program into their computers. These complicated SUA rules appear to be a significant source of quality control errors. For the past few years, the State of Montana has had a waiver in place that substantially simplifies the application of the standard utility allowance. While the Montana approach would be too generous and costly to extend nationally, it indicates that simplifying the application of the standard utility allowance results in significant simplification and a marked reduction in errors related to the shelter deduction. Simplification of the rules governing states' use of the standard utility allowance should be considered.

Another approach that would promote simplification is greater alignment of the food stamp program with Medicaid. The number of households receiving both Medicaid and food stamps now substantially exceeds the number of households either program has in common with TANF. Among food stamp households with children, almost twice as many received Medicaid as received cash assistance in 1999, the most recent year for which these data are available. Individuals who receive both Medicaid and food stamps but do not receive cash welfare benefits are primarily members of working families. This suggests that efforts to better align food stamp and Medicaid rules could both simplify program administration and improve access for eligible working families.

In 1999, about one million low-income children receiving food stamps were uninsured; nearly all of these children were eligible for Medicaid or a state health insurance program for children funded under the SCHIP block grant. Conversely, about three to four million children with incomes under the food stamp income eligibility limit received Medicaid but not food stamps. In both cases, the children enrolled in one program but not the other were largely children from working families. Better coordination between the programs could help to connect more of these eligible nonparticipating children to the programs. Accordingly, efforts should be made to develop simpler joint application forms and procedures by which working families can apply for Medicaid and food stamps together, preferably in settings outside of the welfare office.

Such efforts would be facilitated if states were given the option of aligning the two programs more closely by using common definitions of gross income (so long as all major sources of earned and unearned income are included). This would streamline the food stamp income definition rules, enable states to clean out the minutia, and promote coordination between the two programs. States also could better coordinate eligibility reviews if, as a number of states have recommended, the Food Stamp Program revised its rules regarding food stamp recertification procedures to match those used in Medicaid. It might also be useful to test allowing states to use Medicaid or SCHIP verification procedures in determining food stamp eligibility for families seeking both benefits. (If food stamps and Medicaid were better coordinated, states also could consider building additional work support programs, such as child care, into a basic package that working-poor families could access through "one-stop shopping" and joint applications.)

Also worth testing are strategies to enable working families to apply jointly for food stamps and health insurance in a setting outside the welfare office. Most states now provide avenues for eligible working families to apply for Medicaid or SCHIP benefits outside the welfare office, either for the entire family or for children. As an increasing number of poor families become employed and are able to apply for health insurance without going to the welfare office, the risk increases that they will fail to secure food stamps if they have to go to the welfare office to apply. The Committee could establish pilot projects to test such approaches as co-locating food stamp eligibility workers with Medicaid and SCHIP eligibility workers at hospitals, community health clinics, and other health providers serving large numbers of low-income working families.

II. Providing Adequate Food Purchasing Power

The benefit reductions in the welfare law, which reduced the food purchasing power of all poor households receiving food stamps, were instituted in a time of substantial budget deficits. We are now in an era of surpluses. The adequacy of food stamp benefits should be reexamined as part of the reauthorization process.

S. 583 contains several strong proposals to improve benefits. The following are a few additional options the Committee may wish to consider. (This list is not intended to be exhaustive.)

Restoring Food Stamps' Sensitivity to Inflation -- The food stamp program's standard deduction is supposed to reflect the costs of various necessities that families incur. Because these costs tend to rise with inflation, Congress included an inflation adjustment when it acted on a proposal made by Senator Dole and established the standard deduction in 1977. The welfare law terminated these inflation adjustments, however, in part because some House members were concerned that the Consumer Price Index used to make these adjustments overstated inflation. (The version of the welfare law the Senate Agriculture Committee approved in 1995 did not repeal this inflation adjustment.) Since enactment of the welfare law, the Bureau of Labor Statistics has made a series of improvements in the CPI; as a result, it now rises more slowly. In addition, next year the BLS will unveil a new, alternative CPI with a further important methodological improvement. The alternative CPI, which BLS will be able to issue only once a year rather than monthly, will rise more slowly than the regular CPI. The Committee could consider restoring the inflation adjustment, using the alternative CPI. Alternatively, S. 583 contains an innovative proposal to set the standard deduction at 10 percent of the poverty line, reflecting the fact that if income is that low, virtually none of it may be available to buy food. Such an approach also has another attribute -- it addresses an anomaly in the current standard deduction structure, under which a single individual receives the same standard deduction as a family with children.

Protecting the Elderly and Disabled -- Many elderly and disabled receive small food stamp benefits (such as benefits of $10 or $15 a month) and may choose to let these benefits accrue over a few months before using them in a trip to the grocery store. We are concerned about reports that some elderly and disabled individuals who save up several months of benefits before making a single large shopping trip have had their benefits terminated under the EBT rules that are in use in a number of areas. Current policy in many EBT systems takes benefits off-line after three months. These elderly and disabled recipients may believe mistakenly they no longer are eligible for food stamps because their EBT cards do not work. This issue warrants examination.

III. Gaps in Coverage

Finally, I would urge the Committee to reexamine gaps in coverage that have emerged in the aftermath of the welfare law. I am referring here to the two most severe food stamp provisions of that law -- the denial of eligibility to most low-income legal immigrants and the three-month limit imposed on the receipt of benefits by individuals aged 18-50 who are not raising minor children and are out of work.

In both of these areas, the provisions of the welfare law are considerably harsher than those this Committee adopted when it fashioned its version of the welfare law in 1995. There was a large difference in these areas between the House position and that which this Committee took. Furthermore, the provision in the welfare law that relates to individuals aged 18-50 goes well beyond the provisions that both the Senate and House Agriculture Committees originally adopted; this provision became more harsh as a result of an unexpected last-minute amendment offered on the House floor in July 1996 by several House members who sought to offset the increased costs the final version of the welfare bill contained for child care and a few other items by cutting more deeply into food stamps. The harshness of these provisions, which is uncharacteristic of this committee, warrants reexamination.

In the case of the 18-50 year-old adults, the Committee may want to consider the provision it adopted in 1995. That provision was tough but not draconian.

With regard to immigrants, I would note that the food stamp program's restrictions on legal immigrants are substantially more severe than those that apply in SSI, Medicaid, SCHIP, TANF, or any other means-tested program. The food stamp program is the only means-tested program that denies eligibility to large categories of poor legal immigrants who entered the United States before August 22, 1996, the date the welfare law was signed.

Because legal immigrant parents who entered the country before August 22, 1996 are ineligible, many immigrant families apparently do not understand that their children may qualify for food stamps. As noted earlier, the number of citizen children (with legal immigrant parents) who are receiving food stamps fell by a stunning 74 percent -- or one million children -- between 1994 and 1998.

In addition, in Medicaid and TANF, states have the option of making legal immigrants who entered the country after August 22, 1996 eligible for assistance after they have been in the country for five years. Most states are electing this option. In the State Children's Health Insurance Program (SCHIP), states are required to make such children eligible after they have been here five years. But in the food stamp program, families and children remain ineligible at the five-year point. As noted earlier, S. 583 contains a broad restoration of legal immigrant eligibility.

The confusing food stamp immigrant eligibility provisions now in law -- under which some members of immigrant households may be eligible and others ineligible -- also are a significant source of complexity and error for states. Prior to enactment of these rules, average food stamp error rates were about the same among immigrant and non-immigrant households. Since establishment of the new, complicated rules related to immigrants, error rates among households in which an immigrant resides have risen significantly.

Conclusion

The food stamp programs does a great deal of good and is one of our most valuable and important programs. However, it is not as effective as it can and should be. Reauthorization provides an opportunity to build on the program's strengths and to address the problems that have arisen in it.

1. In 1999, some 87 percent of food recipients were children, employed individuals, elderly or disabled individuals, adults caring for an incapacitated person, or adults who were not employed themselves but were part of a household in which another member worked.

2. This figure measures the change in the number of people who are poor before means-tested benefits are counted.

3. These data on the depth of poverty use a broad measure of poverty that most analysts favor, under which non-cash benefits such as food stamps are counted as income.

4. USDA recently instituted a procedure of making an adjustment in state error rates, when figuring sanctions, to provide some relief to states with caseloads that include a larger-than- average percentage of working households or that recently have experienced a substantial increase in the number of such households that receive food stamps. This adjustment is a useful first step. It is not in statute or regulations, however, and states are not assured of the continuation of this policy.



LOAD-DATE: July 23, 2001