maximum standard utility allowance. The regulations do not provide for prorating the allowance if expenses are shared and, as a result, State and local officials told us that overallowances for utility expenses occur frequently. If both households are food stamp recipients, both can be getting the maximum allowance, which will probably total more than the actual cost of the utilities.

Under the regulations, only households which incur certain utility costs separate and apart from their rent or mortgage payment can use the allowance. Eligibility workers told us that applicants who are not billed for utility services by utility companies separately from their landlords are nonetheless receiving the maximum benefits of the standard utility allowances. For example, the workers explained that to provide the documentation needed to show that they have utility expenses, the applicants provide rent receipts annotated to indicate that a specific portion of their rent is a charge for utility services. Regardless of how small the amount indicated on the rent receipt, or how many other households live at the same address, the eligibility workers said they are required by the regulations to credit the applicant with the maximum allowance.

In the audit of Dade County, Florida that I mentioned earlier, the Department's OIG statistically projected that about 10.3 percent of all the households that received benefits in September 1979 took advantage of this loophole, resulting in overpayments of $138,000 for that month alone. In nearly half of the identified cases, the OIG found that the notations on the recipients' rent receipts were for amounts below the cost of the minimum utility services required by the county. Yet, these recipients received the maximum allowance.

In our report we recommended that FNS revise its regulations to require that States prorate the standard utility allowances in cases where two or more households share the utility expenses of the same dwelling, whether or not all of the households are food stamp recipients.

FNS' management has not provided enough help to States

Our report concluded that FNS had not been very effective in managing the food stamp program because it has not developed meaningful standards to measure and compare States' efficiency and effectiveness in carrying out food stamp legislation. The indicators presently available to FNS are meaningless in comparing States' performances. Although the 1977 Food Stamp Act called for the development of standards to assist in reviewing States' performance, FNS has not yet complied. FNS awarded a contract to a management consulting firm for a performance measurement system which was supposed to be completed in November 1980; however, we did not believe the study would provide a usable system because of deficiencies in the raw data being used by the firm. In a work plan, submitted to FNS in March 1980, the contractor pointed out the problems with the raw data and acknowledged that the study would not result in a usable performance measurement system. The study, however, was not terminated or revised at that time. Some States have begun developing their own performance measurement systems. We recommended that FNS meet with representatives of these and other interested States to develop a performance measurement system that could be used nationwide. Without such a system, we felt that FNS could not perform the basic management function of comparing the operating efficiency and effectiveness of the various States. Nearly 4 years have gone by since the passage of the 1977 Act and FNS still has not met the Act's mandates. Accordingly, we recommended that the Secretary of Agriculture be directed to take immediate steps to develop a usable performance measurement system, to report a timetable for its development to the appropriate congressional committees, and to report periodically on the progress in this area. State officials told us that FNS' delays in issuing regulations placed a great deal of pressure on them to maintain program continuity. They said they were forced to issue internal State operating guidelines and report to FNS on their activities without formal guidance. For example, although FNS' certification manual implementing the 1977 act was still in draft form as late as October 1980, States had to issue manuals to eligibility workers and spent considerable sums of money to train their workers and program their computers in accordance with the manual. Once the Federal manual is completed, States may be faced with the large expense of redoing these activities to conform to the new manual.

Also, most State officials said they had a difficult time getting responses from FNS to their requests for policy clarifications and technical assistance in implementing the 1977 Act. They said that the routing of questions and answers through FNS' regional offices delayed the process and that they could not wait the weeks required to resolve their problems. As a result, they said they usually made the decision and implemented it prior to asking FNS for guidance. Then, if FNS came back with a different interpretation, they revised their policy accordingly.

Careful evaluation of funding needed

In our report, we also said that although we believed there would be significant program savings if our recommendations were implemented, we cautioned that merely reducing the appropriation for the program would not in itself result in improvements. Our recommendations included changes in the basic Food Stamp Act, administrative procedures, and program regulations. Arbitrary reductions in appropriations will trigger sections of the Food Stamp Act that require benefit cutbacks to all recipients-including those who are most in need of assistance-and will not automatically eliminate ineligible persons or reduce program errors or fraud. In fact, carrying out some of the recommendations in our report may require additional administrative funding.

STATEMENT OF DR. ROBERT L. RIZEK, DIRECTOR, CONSUMER NUTRITION CENTER, SCIENCE AND EDUCATION ADMINISTRATION, U.S. DEPARTMENT OF AGRICULTURE Mr. Chairman and members of the committee, the thrifty food plan (TFP), developed by USDA in 1975, is the least costly of the four USDA family food plans. It was adopted as the basis for the coupon allotment for the Food Stamp Program (FSP), effective January 1976 (Federal Register, Vol. 40, No. 231-Monday, December 1, 1975) and incorporated into the Food Stamp Act of 1977, Public Law 95-113, approved 9/29/77.

The TFP is a research-based set of economical and nutritious diets that reflects insofar as possible food choices of households with limited food budgets. The TFP is an assortment of foods that represents as little change from average food consumption of persons in U.S. households with relatively low food costs as was required to provide a nutritious diet while controlling costs. The TFP specifies the amounts of foods of different types (food groups) that households might buy, or obtain from other sources, to provide nutritious meals and snacks for household members. In the plan, amounts of food groups are suggested for men, women, and children of different ages and for pregnant and nursing women (table 1). A plan for any household can be determined by totaling the amounts suggested for persons of the sex and age of household members.

A four-person household—a man, woman, and children 6-8 and 9-11 years of age is used by the Department in setting food stamp allotments. This is not a typical food stamp household. USDA studies have shown that allotments based on this four-person household exceed the sex-age-specific TFP cost for 75 percent of food stamp households.

Households following the TFP may choose from the food groups those foods they enjoy eating. Since the TFP is based on average consumption patterns, not all foods in it are economical. A food list for a family of four for a month (table 2) illustrates the kinds and amounts of foods used in estimating the cost for the plan. In this illustration, the plan provides 23 pounds of red meat and 7 pounds of poultry, despite the fact that chicken is a more economical source of protein than most types of red meat.

Sample menus for a week (table 3) show how foods in the TFP can be combined into appetizing and nutritious meals. The meals are not expected to suit any family entirely. However, they can be prepared from foods typical of those used by survey households operating on low budgets. These foods are generally available in stores across the country. The amounts of foods in the sample menus that are served to family members will differ depending on individual food needs. For example, at a meal, young children may eat 1⁄4 cup of vegetables and teenagers, 1⁄2 cup. Generally, servings of bread and cereals are generous and servings of meat, poultry, and fish are small. Meals for a day for a man following the TFP, on the average, contain about 11⁄2 cups of milk, 4 to 5 ounces of cooked lean meat or alternate such as egg or legumes, 3 to 4 servings of vegetables and fruit, and 7 or more servings of cereal, pasta, and bread and other bakery products.

How TFP was developed

Data from the USDA Household Food Consumption Survey 1965-66, the most recent such data available when the plan was developed, were used to develop the TFP. In developing the plan, minimal changes in average consumption patterns required to meet nutritional and cost constraints were systematically identified by use of a quadratic programing model. Generally, changes were reductions in quantities of meat, poultry, and fish, fruit, and vegetables (other than_potatoes), and increases in quantities of potatoes, dry beans, and grain products. Figure 1 shows how food in the TFP compares with average food consumption patterns for a fourperson household from a more recent survey conducted in spring 1977.

Nutritive value of TFP

The TFP provides for nutritious diets. It was designed in 1975 to provide the Recommended Dietary Allowances (RDA) set in 1974 by the National Academy of Sciences-National Research Council (NAS-NRC) for protein, calcium, iron, magnesium, vitamin A, thiamin, riboflavin, niacin, vitamin B6, vitamin B12, and vitamin C. To meet their iron allowances, young children, teenage girls, and women of childbearing age will need to select frequently foods that provide important amounts of iron, such as liver, heart, kidney, economical lean meats, dry beans, dry peas, darkgreen vegetables, dried fruits, cereals with added iron, and molasses. Such special selections are also necessary for meeting allowances for iron in the more expensive food plans. Levels of vitamin B. and magnesium in the TFP for a few sex-age categories were below the RDA but above levels in usual consumption patterns. Estimates were not made of levels provided by the plan of some nutrients for which there are allowances. Phosphorus levels were not calculated but are known to be well above the allowances in most U.S. diets. The use of iodized salt is recommended as an efficient way to supplement dietary iodine. The requirement for vitamin D for normal persons can be met by exposure to sunlight. However, for infants and elderly persons whose activities limit their exposure to sunlight, the allowance should be provided in the diet by such foods as eggs, liver, butter, and milk fortified with vitamin D or by supplementation. Insufficient reliable information was available on the content in foods of three other nutrients for which allowances were set in 1974-vitamin E, folacin, and zinc-to make reliable estimates of levels provided by the plans.

A TFP developed to meet the RDA would be expected to provide generous amounts of nutrients for most persons. The NAS-NRC states that the basis for the RDA is such that “even if a person habitually consumes less than the recommended amounts of some nutrients, his diet is not necessarily inadequate for those nutrients." Although USDA uses the RDA as a standard in developing food plans, it recognizes that failing to meet the RDA does not imply dietary inadequacy. This is fortunate, because nationwide studies show that very few people-rich or poorhave diets every day that meet the RDA for all nutrients. These studies also show that average calorie intakes for sex-age categories are as much as 25 percent below the RDA for energy; yet obesity continues to be a national problem.

Fat provides 30-39 percent of energy in the TFP, depending on the sex-age category. Sugars, other than found naturally in foods, in the plan provide 12 to 16 percent of energy-less than levels in consumption patterns used as the basis for the plan. Cholesterol levels do not exceed 350 mg per day.

Allowance for food waste

In developing the plan, the RDA were increased by 5 percent to allow for some discard of edible food without jeopardizing the nutritional quality of the diet. (The discard of inedible parts of food, such as excessive fat and drippings from meat and poultry, peelings and bone, and losses of vitamins in cooking, is allowed for in the nutritive values used.) Allowance for edible discard is believed necessary because some edible food is discarded in most homes in the preparation of food, as plate waste, or due to spoilage. Household discard of edible food of 5-11 percent has been estimated from admittedly inconclusive studies of food waste by USDA and the University of Arizona. Our Center is funding a study of methods for measuring household food losses at the University of Arizona.

How TFP costs are estimated

U.S. average costs of foods in the TFP for individuals in each of the 14 sex-age categories and for selected families are estimated by the Consumer Nutrition Center each month and released by USDA. (Costs for the other three plans-low-cost, moderate-cost, and liberal-are also estimated.) Average prices paid for almost 2,000 foods by survey households with relatively low food costs are used as basis for the TFP estimates. Such prices reflect the assortment of food container sizes and brands, quality of food selected, and price levels of stores of purchase for families who use food at relatively low cost. Prices are updated monthly using the Bureau of Labor Statistics' Consumer Price Index for the most detailed food groupings.

The September 1980 cost for foods in the TFP for the four-person household ($233 per month) is the basis for the food stamp allotment for four-person households with no income (effective for the 1981 calendar year, in accordance with the Food Stamp Act Amendments of 1980). For smaller and larger households, the maximum allotment is adjusted for the number of household members and for economies associated with buying and using food in large households relative to small ones.

Many households, whether or not they are eligible for food stamps, do not select nutritious assortments of food. Receipt of enough food stamps to buy the TFP, of course, will not assure that the recipient gets a nutritious diet. Food stamp recipi

ents may not use their stamps to purchase foods as suggested in the TFP. They may make more or less economical and nutritious choices than those in the plan. The TFP is but one of many combinations of foods that could be developed at extremely low cost. Amounts of food groups in consumption patterns could be changed in other ways to provide nutritious diets. While such combinations would deviate further than the thrifty plan from food consumption patterns as reported in 1965-66, they might be as acceptable as the TFP to some households. Other plans at the same or lower cost than the TFP could be developed if selections of foods within food groups were limited to only those foods which are less expensive, rather than selections typical of those in survey households. However, for purposes of estimating the nutritive value and the cost of a plan for use nationwide, average selections of foods in food groups based on those made by survey families with relatively low food costs have been used.

TFP costs and the cost and quality of U.S. diets

The Nationwide Food Consumption Survey 1977-78 provides information that shows that some households spend at levels below the maximum food stamp allotment level based on the cost of the TFP. It also shows that some households spending at the maximun allotment level used food that provides the RDA for 11 nutrients studied.

The TFP was developed at a cost level below which 10 percent of U.S. households spent in 1965-66. In 1977-78, of over 14,000 U.S. households surveyed, 11.6 percent used food at home during the survey week valued at less than the maximum food stamp allotment for a household of that size.

The nutritive value of household diets was estimated and compared to the 1974 RDA for 11 nutrients. (This is but one of several measures of nutritional quality of household diets that might be used.) Of households using food valued at the full food stamp allotment level, 9 percent used food that met the RDA for 11 nutrients and 33 percent used food that met 80 percent or more of the RDA for 11 nutrients, as shown in Figure 2. At a cost level 15 percent above the allotment level, 18 percent of the household diets met the full RDA and 51 percent met 80 percent of the RDA or more. At a cost level 15 percent below the allotment level, 3 percent of the household diets met the full RDA and 16 percent met 80 percent RDA or more. As discussed earlier, diets that meet the RDA would be expected to provide generous amounts of nutrients for most persons; and failure to meet the RDA for all nutrients does not imply dietary inadequacy.

Table 4 shows the average money value of food per person per week, for five groups of households, by income reported in the spring of 1977. The three lowest income groups had money values, on the average, that were about the same. It appears that differences in food costs among households within an income group are larger than differences among income groups, and that food costs may be related less to household income than to other factors. Factors affecting the cost and nurtritional quality of diets of low-income households are being studied under contract with the University of Missouri.

Since the TFP was developed in 1975, new information has become available on food consumption and food prices of households (Nationwide Food Consumption Survey 1977-78), on nutritional goals (1980 Recommended Dietary Allowances and other dietary standards), and on the content of several additional nutrients in an increasing number of foods. USDA is now using this new information to evaluate and revise the thrifty food plan as well as the other more costly plans.

TABLE 1.-THRIFTY FOOD PLAN; AMOUNTS OF FOOD FOR A WEEK

s Weight in terms of dry beans and peas, shelled nuts, and peanut butter. Count 1 pound of canned dry beans, such as pork and beans and kidney beans, as 1/3 pound.

6 Cereal fortified with iron is recommended.

7 For pregnant and nursing teenagers, 7 quarts is recommended.