B-227682.2, Aug 16, 1989, 68 Comp.Gen. 618

B-227682.2, Aug 16, 1989, 68 Comp.Gen. 618

Procurement - Payment/Discharge - Utility services - Payment procedures - Administrative policies - Revision The General Accounting Office has no objection to a General Services Administration (GSA) modified proposal to combine elements of fast pay procedures and statistical sampling techniques to pay and audit utility invoices. GSA's modified proposal is a valid sampling plan because it is designed and documented to provide for effective monitoring, a sampling of those invoices not subject to complete audit coverage, audit emphasis commensurate with the risk to the government, and a basis for the certification of payments.

General Services Administration's Modifies Proposal for Post- Payment Examination of Utility Invoices by Statistical Sampling:

This advance decision to the General Services Administration (GSA) responds to a request from Raymond A. Fontaine, GSA Comptroller, for our approval under 31 U.S.C. Sec. 3521 (1982) of a modified proposal to combine elements of fast pay procedures and statistical sampling techniques to pay and audit utility invoices. In 67 Comp.Gen. 194 (1987), we did not approve an earlier proposal submitted by GSA because the sampling plan then under consideration did not meet certain requirements for our approving such a plan. Our decision discussed alternative modifications to GSA's proposal which could result in a valid statistical sampling program and invited GSA to submit a modified proposal for our further consideration. 67 Comp.Gen. 194, 200-201. For the reasons discussed below, we have no objections to GSA's modified proposal.

BACKGROUND

As an alternative to the current post-payment examination of all invoices, GSA initially proposed to audit only invoices which exceeded GSA's estimates of utility costs for the corresponding month. GSA further proposed to establish monthly estimates that are derived from historical data accumulated over the last 2 years in conjunction with estimates of current usage (accruals). To help protect against the estimate being overstated, GSA proposed a monthly sampling of 1 percent of payments made to the vendors on record for the prior 3 months. However, this sample only was for the purpose of determining the accuracy of GSA's estimating of billings, not to determine whether the invoices themselves were accurate. If GSA's estimate exceeded the invoices by 15 percent it would then make adjustments to its future estimates of billings. Other procedures also were proposed to supplement this plan. 67 Comp.Gen. at 196.

We did not approve the plan initially proposed by GSA. Several factors contributed to our view that the proposed plan did not provide adequate controls to reasonably assure that losses occurring from potential overbillings would not exceed the savings from the proposed sampling plan, and therefore did not adequately protect the interests of the Government. First, the proposed plan did not provide for a sampling of all invoices since it excluded from the sample all invoices falling below GSA's estimate. 67 Comp.Gen. at 199-200. Second, GSA did not identify the specific savings that would be effected through use of the proposed plan. 67 Comp.Gen. at 199, n. 10. Finally, it did not provide monitoring to assure that all audit results are captured and analyzed in a way that reaffirms the reliability of, or identifies and corrects problems with, the sample audit approach or estimating procedures. 67 Comp.Gen. at 201.

GSA's Modified Proposal

The modified proposal submitted by GSA proposes to establish the accrual process described in GSA's original request. However, as a means of monitoring the effectiveness of GSA's accruing of utility charges, GSA plans to audit all invoices that fall outside a 15 percent tolerance of that accrual. Further, it also proposes a monthly random sample audit of 3 percent of the remaining paid invoices.

To assist it in determining the appropriate meaningful sample, GSA's Office of Finance employed the services of a professional statistician and forwarded a copy of the statistician's report entitled "Random Sample Size for Utility Bill Audits" for our review. The report indicates that initially the net savings due to employment of the 3 percent monthly random sample of utility invoices should be about $97,000. /1/ Also, there is a high probability that all vendor's billing will be selected within a reasonable period of time as part of the 3 percent sample. GSA will monitor the effectiveness of the sample size on an annual basis to determine if adjustments are required.

Added review and analysis for utility bills will be forthcoming upon completion of Public Building Service (PBS), National Energy Usage Analysis System Center (NEUASC) located adjacent to the National Payment Center in Fort Worth, Texas. The NEUASC will consolidate certain PBS utilities and fuels activities into one location. Functions of the NEUASC will include: ensuring the integrity of utility invoices by evaluating charges levied; checking invoices against forecasted building usage and cost; and identifying and resolving discrepancies in invoices and GSA estimates.

GSA also proposes to establish internal controls to effectively monitor errors detected within the post certification sampling process to insure undetected billing errors are minimized. The primary control will be to audit all bills paid in the past 12 months for an account whose sampled bill is proven erroneous. Additionally, the vendor number of the utility, along with the amount and type of error, will be kept on file for 5 years for annual review by GSA's Accounts Payable to insure no pattern of abuse is being established. If a vendor has established a pattern of abuse, all payments to that vendor will receive verification.

Finally, the vendors involved are utilities with which GSA maintains a continuing business relationship barring some unforeseen unusual occurrence. Thus, once overbillings are discovered, GSA will have ample opportunity to recover these amounts by setoff against future billings. This is an important consideration when assessing the overall risk to the government under GSA's proposal.

ANALYSIS

Prepayment audits are generally required by GAO, Policy and Procedures Manual for Guidance of Federal Agencies, tit. 7, Sec. 19 (TS 7-41 Jan. 18, 1985). In order for this Office to find that an agency's proposal to combine elements of fast pay procedures and post-payment statistical sampling techniques may be used in lieu of the standard prepayment audit (including statistical sampling), the agency's proposal must adequately protect the interests of the government. /2/ Therefore, the agency's proposal should at a minimum provide for: /3/

1. A statistical sampling of all invoices not subjected to complete audit coverage.

2. A commitment to monitor and modify the sampling program on the basis of results of the actual operation of the plan and other related developments, and to take responsive action when weaknesses or errors are detected by the sampling process.

3. A probable net saving due to the use of the statistical sampling technique.

GSA's modified proposal, unlike its initial proposal, provides for a sample of those invoices not subject to complete audit coverage based on assumptions set forth in the Report prepared by GSA's statistician. GSA proposes a monthly random sample of 3 percent of the paid invoices. addition, it proposes to audit all invoices that fall outside a 15 percent tolerance of GSA's accruals. Thus, the proposed plan provides for limited audit coverage of paid invoices falling within the 15 percent tolerance and complete audit coverage of all invoices that fall outside the tolerance. This seems reasonable, at least initially, based on the 5 percent error rate assumed for establishing the plan.

The modified proposal also includes monitoring of the proposed sampling plan to determine its effectiveness and to permit GSA to make timely modifications to the sampling plan to correct any deficiencies identified through its actual operation. The other reviews and analysis to be performed by GSA will provide additional safeguards to assist GSA in monitoring the effectiveness of its sampling plan and to identify problem utilities requiring special attention.

The materials GSA provided also identify the net savings to the government expected initially from use of the sampling plan. While admittedly these savings are based on certain assumptions (for example that the error rate will not exceed 5 percent), they provide a basis for making an initial determination as to whether to implement the plan. For example, based on past history, GSA expects error rates to run closer to 3 percent. Furthermore, the assumption recognizes possible costs through losses in overbillings and compares them to potential savings resulting from decreased audits. While the cost of collecting overpayments in some situations also must be factored into this equation, it is not important to GSA's modified proposal. The vendors involved here are utilities with whom GSA maintains a continuing business relationship. Therefore, once overbillings are discovered, GSA will have ample opportunity to recover these amounts by setoff against future billings without incurring additional costs.

For the reasons stated above, GSA's proposal provides for a meaningful, valid statistical sampling of invoices, an adequate monitoring program, and probable savings to the government. Therefore, we have no objection to GSA's modified proposal to combine elements of fast pay and procedures statistical sampling techniques to pay and audit utility invoices since, if implemented properly, it should adequately protect the interests of the government.

/1/ This is based on an initial assumption of a 5 percent error rate in billings. Once data is gathered through sampling of invoices, GSA can determine whether its assumptions are correct and whether it is necessary to adjust the sample to protect the government's interest, or to revert to a post-payment audit of all vouchers because statistical sampling proves not to be cost effective.

/2/ We address only whether GSA's proposal complies with our standards. Compliance with our standards does not relieve an agency from the need to satisfy any other applicable requirements, such as those established by the Office of Management and Budget or the Federal Acquisition Regulation.

/3/ As we stated in our previous decision, we are willing to entertain alternatives to our prepayment or 100 percent post-payment audit requirement in appropriate circumstances. 67 Comp.Gen. at 198, 199. However, when such alternatives are approved, the agency should periodically evaluate whether advances in communications technology as well as other electronic systems capabilities afford the agency the opportunity to conduct prepayment or post-payment audits of invoices and assure timely payments to vendors in an economically feasible manner.