Each State's sample cases are reviewed by State QC reviewers to determine the accuracy of the grant amount and to verify the recipient's eligibility in a given month. For each case, the reviewer verifies factors, such as family income, resources, and other basic program requirements, which affect both eligibility and the grant amount. The reviewer verifies these factors by contacting the recipient and often other (collateral) sources, such as landlords, employers, and banks.

Each State compiles the results of its review and computes both case and payment error rates for three categories of error: payments to ineligible recipients, overpayments to eligible recipients, and underpayments to eligible recipients. 1/

HHS then selects a subsample from cases in each State's QC sample and re-reviews the cases. The purpose of the rereview is to assure that the States are conducting their QC reviews correctly and to validate the State error rate. The results of the Federal re-review and the State review are combined using a statistical formula to compute the official State error rate.

The

Each State develops a corrective action plan, based on the payment errors identified to deal with their causes. plan might include provisions for staff training, program procedural revisions, or changes in AFDC recipient eligibility requirements, as appropriate to the cause.

FISCAL SANCTIONS

The 1973 HHS regulations that established the current QC program provided for the assessment of financial sanctions or penalties against States with high error rates. According to the regulations, the Federal share of total erroneous AFDC expenditures in excess of 3 percent to ineligible cases and 5 percent for overpaid eligible cases in each State was to be withheld. States were to reduce their error rates below the 3- and 5-percent tolerance levels over three successive QC sampling periods.

1/QC also reviews a sample of negative case actions which include denials of aid and aid terminations. We did not review any negative case actions.

Fourteen States challenged the legality of the sanction regulation in court. The court ruled that HHS could impose sanctions, but that the 3- and 5-percent limits were arbitrary and capricious and, therefore, unenforceable.

On July 7, 1978, HHS proposed a new sanction regulation. Federal funds were to be withheld for erroneous expenditures in excess of 4 percent of each State's total benefit payments. States were not required to meet the 4-percent rate immediately. They were only to be sanctioned if they were above the median error rate for all States and had not reduced their payment error rate from the previous QC period by 18 percent. The 18-percent figure represented the national error rate reduction that had been achieved between the April-September 1973 and July-December 1976 sample periods.

In response to opposition toward the sanctions proposed in 1978, HHS issued a final regulation in March 1979 that contained less severe sanctioning criteria. The regulation reduced the error improvement rate to 6.4 percent, which represented the national error rate reduction achieved between the January-June 1976 and July-December 1977 sample periods, eliminated the 4-percent goal and committed HHS to conduct a study to determine what the ultimate error rate goal should be.

In its report on deliberations on a fiscal year 1979 supplemental appropriations bill, the House-Senate conferees directed that HHS issue regulations requiring States to reduce the AFDC payment error rate to 4 percent by September 1982 or lose Federal matching funds associated with erroneous payments in excess of the target. In January 1980, HHS issued final rules to implement the sanctions directive.

FISCAL INCENTIVES

There currently is a fiscal incentive for the States to reduce their AFDC payment error rates. In the Social Security Amendments of 1977 (Public Law 95-216), the Congress established a formula by which States that reduce their QC payment error rates below 4 percent can participate increasingly in the Federal share of the money saved. For each one-half percent below 4 percent, a State receives an additional 10 percent of the Federal funds saved until its error rate is reduced below 2 percent, when the State's

maximum share of the Federal funds saved is 50 percent. This provision became effective during calendar year 1978.

SCOPE OF REVIEW

We made our review at HHS headquarters in Washington, D.C.; HHS regional offices in Boston, Chicago, New York City, Philadelphia, and San Francisco; and in California, Hawaii, Indiana, Maine, Maryland, and New York. These States were selected based on the size of their AFDC caseload and their payment error rates.

We also interviewed Federal and State officials to obtain their views on the use of fiscal sanctions by the Federal Government.

We did not review the overall accuracy of the individual AFDC case reviews made by State and Federal QC reviewers. Instead, we concentrated on identifying overall system problems that reduce the effectiveness of the QC system.

We

We analyzed the six State programs to identify differences which could affect the error rates of each State. examined about 3,300 QC sample cases from the January-June 1978 sample period in five of the States to determine the type and amount of management information that was not being gathered by the QC system.

We reviewed the QC sampling plans for selecting AFDC cases for review in the six States where we made this audit. In our opinion, all of the sampling plans yielded a representative sample of active AFDC cases.

We provided copies of the draft report to HHS and the States and obtained oral comments from HHS and written comments from the six States. (See app. I - VI.)

CHAPTER 2

FISCAL SANCTIONS SHOULD

BE DISCONTINUED

To encourage States to improve their AFDC program administration, HHS and a congressional conference committee have taken steps to impose fiscal sanctions, which will result in Federal funds being withheld from States that have rates of erroneous payments that exceed error rate tolerances based on the QC system.

In our opinion, fiscal sanctions place the Federal and State governments in an adversary role at a time when cooperation is needed for reducing errors. Because a high error rate can result in sanctions, States have an incentive to identify fewer errors. As a result, basing sanctions on QC measurements can discourage error identification and thereby reduce QC usefulness as a management tool for developing corrective actions aimed at the causes of errors.

In addition, we question whether current QC system error rates are sufficiently comparable to be the basis for sanctions because they (1) vary in statistical precision, (2) do not clearly show the effects of program and policy differences among the States, and (3) are not based on the same error identification procedures in all States. (See ch. 3.)

PAST EFFECT OF SANCTIONS

HHS officials responsible for QC in the AFDC program believe that fiscal sanctions or the threat of them has been a major impetus to States in reducing their AFDC error rates. They pointed out that, from the time sanctions were first proposed until the error rate basis for sanctions was invalidated by the courts, error rates declined significantly.

While the threat of sanctions in the past was undoubtedly a factor which helped focus management attention on reducing errors, we believe other factors, such as increased public awareness of the existence of high error rates in the AFDC program, also provided an impetus to States to improve program administration. As discussed later in this report, we believe that applying sanctions based on QC results will not treat the States equitably.

We generally favor incentives to encourage error reduction and preserve Federal-State cooperation, as indicated in recent testimony on proposed amendments to the Food Stamp Act of 1977 1/ and in our report, "Review of the Better Jobs and Income Bill" (HRD-78-110, May 23, 1978).

CONCERNS ABOUT NEGATIVE EFFECT

OF SANCTIONS ON COOPERATIVE STATE-
FEDERAL EFFORT TO REDUCE ERRORS

Cooperation between the Federal Government, which establishes the basic framework for AFDC, and the States, which administer the program, is in our opinion essential for effective error reduction. The Federal Government possesses a nationwide perspective on program administration. States, however, are responsible for day-to-day operations of AFDC and are better suited for identifying the causes of errors.

An example of a cooperative effort is the development of profiles on error-prone cases. HHS has studied characteristics of cases that are predictive of errors and has developed profiles of error-prone cases. Such profiles have proven successful in targeting error reduction efforts on these cases. HHS cannot, however, do a profile analysis on a State's AFDC caseload unless the State provides it with information on its error cases. According to HHS officials, States are becoming more aware of error-prone case profiles and are requesting them for their AFDC caseloads. However, because of the negative feelings generated by sanctions, some States may be reluctant to cooperate in this effort.

Several States have pointed out to HHS that sanctions could harm the Federal-State relationship in administering the AFDC program. For example, in a September 1, 1978, letter to the Secretary of HHS, the Secretary of Maryland's Department of Human Resources stated that:

"*** the debate regarding the uses of the
Quality Control system has overshadowed any
progressive efforts and good will that could

1/Testimony on proposed amendments to the Food Stamp Act of 1977 (H.R. 4318), October 17, 1979, before the Subcommittee on Domestic Marketing, Consumer Relations, and Nutrition, House Committee on Agriculture.