percent behind the cost of the Thrifty Plan. This further invalidates the assertions that poor school children are receiving too much food.

AFDC AND FOOD STAMP BENEFITS AS PERCENTAGE OF POVERTY LEVEL UNDER CURRENT BENEFIT RULES AND UNDER ADMINISTRATION'S PROPOSED CHANGES

Note: Percent of poverty was calculated as total monthly disposable income divided by $7,070, the poverty level for a nonfarm family of 3 as published in the Federal Register, Mar. 5, 1981, to take effect Apr. 6, 1981.

Eliminating the adjustment on the maximum shelter deduction and the standard deduction

The Administration's food stamp proposals contain other provisions that would also harm many families below the poverty line. If a family's income fails to rise with inflation (as is the case for many AFDC families, for example), and the family's heating bills increase sharply, the family has less disposable income to spend on food. The current Food Stamp Program recognizes this and adjusts food stamp benefits accordingly. Under one of the Administration's proposals, however, this adjustment would no longer be made for many families.

In figuring a household's disposable income, an excess shelter deduction is now provided for the amount by which shelter costs exceed 50 percent of the household's net income, up to a maximum of $115 a month. This maximum is annually adjusted to reflect changes in shelter and utility costs.

The Administration is proposing to end this adjustment, despite the fact that utility companies are predicting steep rises in home heating costs next year (D.C. area companies have predicted an 80 percent rise if natural gas deregulation is accelerated). Under the Administration's proposal, many households with increased shelter costs and decreased funds available for purchasing food-would be allowed no adjustment in food stamp benefits to reflect their loss of disposable income. This proposal would be particularly harsh on families in northern parts of the country. Half of all food stamp households in New England and 25 percent of all households in the nation as a whole already are at the maximum shelter deduction and would be hit by this cut. It should be noted that the current shelter deduction is the only feature in the Food Stamp Program that allows for any regional variances in the cost-of-living to be taken into account.

The cutback in the shelter deduction should be viewed not only alongside proposals to speed deregulation of oil and natural gas, but also alongside proposals to fold low-income energy assistance into a bloc grant that is funded 25 percent below last year's level.

While I believe that the proposal to eliminate the adjustment of deduction levels should not be adopted, some change in how the adjustments are made may be in order. In adjusting the maximum shelter deduction, the home ownership component of the CPI could be excluded. Most food stamp households do not own homes and consideration of home ownership costs is not needed. This change should provide some savings.

Retrospective accounting/monthly reporting

The Administration's proposal to mandate that all states and counties must institute a retrospective accounting/monthly reporting system also warrants examination. The 1980 Food Stamp Amendments provide for a State option to use retrospective accounting and monthly reporting. The State option was provided because many states and local areas-especially rural areas with small caseloads-do not have the computer capabilities to manage this very complex and highly sophisticated system. Some such states are convinced that this system would not be cost effective and would be extremely difficult for them to manage. This Administration has repeatedly stated that it wishes to remove federal requirements being imposed on states and provide more state flexibility. The proposal not to allow a state option regarding retrospective accounting/monthly reporting, and instead to impose-over the objections of a number of states-federal requirements mandating that all states and counties implement this system, is inconsistent with the Administration's professed philosophy regarding federal-state relations.

Moreover, the Administration's own cost estimates reflect the difficulty and cost of forcing this system on all states nationwide. The Administration is predicting that two years will be necessary to implement this system and that no net savings will be produced by this provision in either fiscal year 1982 or fiscal year 1983. The predicted savings do not start until fiscal year 1984, and the savings predictions for that year are probably too optimistic.

Retrospective accounting and monthly reporting have been tested in several areas in the AFDC program. Basically the system works as follows: households are mailed a form at the end of each month on which they report on the income they received that month. This report is then used to determine their benefit level for a subsequent month. For example, a form will be mailed to recipients at the end of March on which recipients must report their March income. The form will be returned and processed during April. The household's May benefits will then be based on this report.

The system uses retrospective accounting in the sense that May benefits are based on March income. The system uses monthly reporting in that households must file a report each month and those who do not get the form back get cut off the program. Running this system is not easy. Forms must be promptly mailed by the tens or hundreds of thousands each month, and must be promptly processed in the same numbers. Without highly sophisticated and expensive computerization, the system is virtually inoperable.

The major attraction of the system has been thought to be its potential to reduce errors through requiring more prompt reporting of changes in income. Its drawbacks are that it adds complexity and administrative costs, that benefits do not reflect a family's current needs, and that some needy households lose benefits because they have difficulty with the government forms. The AFDC experiments indicate that of those who are cut off because they do not return the form, onefourth to one-half are persons who are actually eligible but who had trouble with the forms.

Last year, Congress established an error rate sanction system that penalizes states with high error rates that fail to reduce them sufficiently, while rewarding states with low error rates. States should be given some freedom in determining how to get their error rates down. It is not sound policy to compel a state that already has a low error rate-and that believes that retrospective accounting/ monthly reporting will not be cost-effective-to spend considerable administrative cost and effort installing this system. As a result, it is sounder policy to maintain a state option in those areas, while continuing to hold states accountable for reduced error rates.

Three final points regarding retrospective accounting also deserve consideration: (1) The idea of retrospective accounting/monthly reporting has always been to try to reduce errors through the filing of monthly reports. It has never been the goal of this system to make families in need wait 30 days before they can begin receiving assistance. In all areas where this system is now in use, households are initially certified on current need, and then are switched to a retrospective system after they are on the program. Last year's Food Stamp Amendments follow this approach. In addition, Secretary Schweiker has testified that this will continue to be the approach followed in the AFDC program. The Committee should take care to assure that a waiting period is not imposed. If a family is in immediate need because its wage-earner has been laid off, or because a husband has deserted or cut off child support, the family should not have to go without food stamps in the meantime. (2) There is no need to require that households composed entirely of elderly and disabled persons who have no earnings must file a report each month as a condition of eligibility. The Administration is not proposing to require monthly reports in the

SSI program, and there is no reason to require it of elderly or disabled persons on fixed incomes who receive food stamps. No retrospective accounting/monthly reporting experiment that includes the elderly and disabled has ever been run. These people's incomes generally do not change from month to month. There is really no good reason to add the administrative expense of processing reports from these households every month, nor is there justification for imposing a major new paperwork burden on these citizens. If such a requirement is imposed, some of the most needy elderly and disabled persons are likely to lose their food stamps because they cannot complete and file the report each month.

(3) Several states are now using monthly reporting in conjunction with prospective accounting in their AFDC programs, and find it just as useful as retrospective accounting in reducing errors. This makes sense, because it is the monthly reports and not the retrospective accounting that produces whatever savings this system generates. States should be provided the flexibility to combine monthly reporting with prospective accounting if they so choose.

Puerto Rico block grant proposal

The Administration is proposing to consolidate all food assistance programs in Puerto Rico into a block grant, to be funded at 75 percent of the current services level. This will cut food assistance programs in Puerto Rico by $300 million, from $1.2 billion to $900 million. The primary programs affected are food stamps, school food programs, and the special supplemental food program for Women, Infants, and Children (WIC).

Nearly all the reductions will have to be taken out of the Puerto Rico Food Stamp Program, however. The WIC program is very small in Puerto Rico (about $20 million a year) and cannot be cut to any significant degree (especially in light of the island's proverty and its infant mortality rate). Little can be cut from school food programs because there are few middle-income children who can be charged more for school meals.

The current Food Stamp Program is large in Puerto Rico. This is because of several factors: there is considerable poverty and a very high unemployment rate (close to 20 percent; and public assistance programs provide very limited benefits, disproportionate to the levels of poverty and unemployment. Because of these factors, a large proportion of Puerto Ricans are eligible for food stamps.

The Administration's block grant proposal would lead to major food stamp cutbacks. Lowering the income limits in Puerto Rico cannot produce the specified amount of savings-too many Puerto Ricans are at the bottom of the income scale. In order to realize the requisite savings, the block grant proposal would very likely lead to across-the-board benefit cuts in Puerto Rico that would affect all recipients, including the poorest. Yet food stamp allotment levels in Puerto Rico are already lower than on the mainland despite the fact that food prices in Puerto Rico are considerably higher than on the mainland. (The prices are higher because most food is imported from the U.S., and shipping costs are built into the final retail price.) Finally, the proposal apparently provides no lead-time for planning or preparing for a block grant. It would be effective six months from now, at the start of fiscal year 1982. It is simply not possible for Puerto Rico to plan for and make the changes needed to lop off 25 percent of its food programs by October 1. This is not a responsible proposal from a management standpoint. An attempt to implement a block grant in October 1981 would likely lead to very grave administrative problems.

Food stamp "cap"

Finally, I should also like to address the matter of the food stamp "cap”.

Nothing plagues the management of the Food Stamp Program more than constant funding crises followed by large numbers of additional changes in the program each time Congress raised the cap. As long as the program keeps changing at the rate it has the past few years, state and local administrators will have a difficult time improving management and reducing errors. Several times last fall, former USDA Inspector General Thomas McBride said that stability, and a respite from constant changes, was basic to improve food stamp management.

Something needs to be done this year to address the problem of the cap. We cannot live with four more years like the past few. Unfortunately, we already know that the Administration's estimates for food stamp spending over the next four years do not accurately reflect the impact of the Administration's budget proposals. The estimates do not take into account the fact that Administration proposals to cut AFDC, unemployment insurance, public service jobs, and trade adjustment assistance will result in increased food stamp costs. More persons will be made eligible for food stamps because of income losses, and some of those currently eligible will qualify for larger benefits. For example, the HHS Department has officially stated

that its proposals to cut AFDC benefits by $1.2 billion will result in $266 million of added food stamp cuts.

In addition, the economic assumptions used by the Administration for both food prices and unemployment are remarkably rosy. Adjustments must be made for the impact on food stamp costs of budget cuts in other programs, and for the possibility that the economy could follow the path predicted by CBO or private forecasters rather than that predicted by the Administration. If new caps are established, but these factors are not taken into account, and annual food stamp crises for each of the next four years—and continued administrative dislocations in the program-will be virtually guaranteed.

STATEMENT OF JOHN T. DEMPSEY, PH. D., DIRECTOR, MICHIGAN DEPARTMENT OF SOCIAL SERVICES AND CHAIRMAN, NATIONAL COUNCIL OF STATE PUBLIC WELFARE ADMINISTRATORS OF THE AMERICAN PUBLIC WELFARE ASSOCIATION

Mr. Chairman and members of the Subcommittee, my name is John T. Dempsey, and I am director of the Michigan Department of Social Services. I also serve as Chairman of the National Council of State Public Welfare Administrators (NCSPWA) of the American Public Welfare Association.

I very much appreciate this opportunity to testify on behalf of the state public welfare agencies regarding the importance of the Food Stamp Program and of the need for improvements and cost controls in this vital component of our nation's welfare system.

The NCSPWA is composed of those officials of each state, the territories, and the District of Columbia, who are responsible for administration of publicly funded income maintenance and social service programs. These programs-which include, among others, Food Stamps, Aid to Families with Dependent Children (AFDC), and General Assistance-provide help to millions of vulnerable low-income individuals and families.

Because it is our daily responsibility to wrestle with state budgetary constraints, staff shortages, complex policies, and the very real needs of the poor and disadvantaged people who come to our agencies for assistance, we feel especially well qualified to speak about the strengths and weaknesses of these programs. We respectfully offer our observations and recommendations in the hope that you will find them useful as you consider reauthorization of the Food Stamp Program.

STRENGTHS OF THE FOOD STAMP PROGRAM

While the Council believes improvements are needed in the Food Stamp Program, our members also see strengths in this national approach to food assistance for the poor that should be carefully examined and preserved.

For example, numerous studies and simple observation tell us that severe hunger and malnutrition have been virtually eliminated in this country in the last 10 to 15 years. There is also considerable evidence that the general health of the nation's poor has improved during this period, and infant mortality rates and other indicators of poverty and malnutrition have declined.

As administrators of programs serving the poor, we can see that expansion of the Food Stamp Program has contributed significantly to these gains. We believe the program's importance as a basic component of the welfare system should not be undermined in efforts to reduce costs, and we urge the Congress to particularly safeguard two of its strongest characteristics. These are the fact that:

(1) Food Stamps are available to the eligible working poor, intact families, and individuals who do not qualify for other types of public assistance or income security programs.

Eleven percent of food stamp households are headed by adults who work full time and still do not earn enough to pull above the povery line. In many states, food stamps are the only form of help offered to families with both parents present and unemployed. They are often the only assistance available to childless couples and individuals who are too old, unskilled, or ill to compete in the job market and too young for other benefits. The Food Stamp Program fills gaps in the present welfare system because it is the only nationwide, noncategorical program. Harsh limitations on eligibility would cause serious hardship for many people and would place crippling strains on the emergency resources of state and local governments and private charities.

(2) Food stamp benefits protect the nation's poorest people against the most devastating effects of inflation and unemployment.

The average annual income of food stamp households presently is about $3,900 a year, well below the poverty level. While all of us have felt the effects of inflation in

79-755 0 - 81 - 26

recent years, food stamp households have been especially hard hit. Between 1975 and 1979, the Consumer Price Index for necessities went up 32 percent, and overall income went up 40 percent. But the income of food stamp recipients increased by only 17 percent.

One reason for this lag in incomes for food stamp households is the fact that onethird of them receive AFDC. Most states do not index AFDC benefits for inflation, and some because of economic forces beyond their control, have actually reduced AFDC payments. Thus, food stamps have provided an increasing proportion of the benefits paid to the poor, particularly in those states suffering the worst effects of unemployment and increased housing and energy costs.

Food stamp benefits presently average only 44 cents per person per meal, or $1.32 per person each day. But these modest benefits often mean the difference between mere subsistence and a miminally adequate diet. Again, efforts to reduce food stamp costs must take into account the effects of inflation and unemployment-not only on the food-buying power of this low-income population, but also on the ability of states to keep up with continually mounting costs that are beyond our control.

One of the lessons coming out of the economic recessions and other national emergencies of recent years is that the Food Stamp Program responds quickly to changing conditions. It has clearly helped to ease the pain of economic decline for hard-hit states and localities. It is also an efficient way to combine the resources of government and the marketplace to provide safety net protection for the leastadvantaged members of our society. These are strengths that should not be overlooked in considering the program's future.

NEED FOR PROGRAM STABILITY

While the Council supports the basic concepts on which the Food Stamp Program is built, we also recognize the need to tighten eligibility and cut costs. We believe the reauthorization process offers the opportunity to examine alternatives. At the same time, we would caution against administratively costly and unmanageable proposals and further upheaval in the state and local administration of this program. Let me explain the concerns of state administrators, based on experience and a genuine interest in the wise use of tax dollars.

In the last three years, three major food stamp bills have been enacted. Each has significantly affected the operation and design of the program. Most major federal regulations have been rewritten once or twice to reflect adjustments in the law, and it has been necessary for states and local agencies to implement hundreds of separate policies and procedures. This has been a costly business in terms of dollars, errors, training time and expense, and disruption of agency operations, and it has affected the overall integrity and direction of the program.

The Food Stamp Program must be permitted to stabilize. It is crucial that we have time to evaluate the impact of changes already made by Congress before considering significant new proposals that may further complicate administration. If we are to have that period of stability, it will be necessary for Congress to remove the authorized cap on program expenditures. The cap on annual spending contained in the 1977 Food Stamp Act has, in effect, opened the door to annual reauthorization of the program. Congress must lift the cap each year to avoid a cut off or reduction of benefits-and we are grateful that you have acted in the last two years to provide funds as needed. But each time Congress acts to lift the cap, new amendments aimed at tightening eligibility and management are also enacted. The cumulative effects of these annual amendments have been a serious drain on state management capabilities. We cannot continue to assimilate major changes, and we ask your help in bringing this situation under control.

We believe there is one other compelling reason for deleting the cap from the law. That is the current unreliability of economic predictions. Costs for the Food Stamp Program are determined primarily by economic conditions-people become eligible and apply for food stamps because they lose their jobs or can no longer stretch low incomes. According to Department of Agriculture figures, more than a million people are added to this program each time the unemployment rate goes up 1 percent; and food stamp costs increase by $150 million a year each time food prices go up 1 percent.

State administrators understand that it is imperative for Congress to find ways to control food stamp expenditures. However, we believe the program can be operated most efficiently if the cap on expenditures is deleted from the law to allow genuine need to be met. Since unemployment and inflation cannot be accurately predicted over any period of time, program spending should not be limited on the basis of current estimates but should be subject to the prevailing economic conditions.