Peggy E. (Peg) Gustafson was sworn in as SBA Inspector General on October 2, 2009. Ms. Gustafson previously served as General Counsel to Senator Claire McCaskill (D-MO), where she advised the...
Report 9-10 – The Small Business Administration’s Fiscal Year 2007 Improper Payment Rate for the Disaster Loan Program
This report summarizes the results of our audit of the Small Business Administration’s (SBA) Fiscal Year (FY) 2007 Improper Payment Rate for the Disaster Loan Program. The Improper Payments Information Act (IPIA) of 2002 requires Federal agencies to review their programs and activities annually, identify programs that may be susceptible to significant improper payments, estimate amounts improperly paid, and report on the amounts of improper payments and actions to reduce them.
Since implementation of the Act, SBA has reported a low improper payment rate for the Disaster Loan Program. However, because several recent OIG audits uncovered a significant number of improper disaster loan disbursements, we initiated the audit to determine whether SBA was properly identifying improper disaster loan payments and correctly estimating the improper rate for the program. We also planned to evaluate whether SBA had developed a plan to reduce improper payments, which is required when an agency is determined to have significant improper payments. However, we eliminated this audit objective as SBA had not developed a plan because it estimated the level of improper payments in the Disaster Loan program to be insignificant.
To assess the accuracy of SBA’s FY 2007 improper payment review process, we compared the Agency’s review methodology with Office of Management and Budget (OMB) guidelines contained in Appendix C of OMB Circular A-123.
These guidelines, which were published in 2006, provide agencies with specific instructions for estimating and reporting the rate of improper payments. We also consulted OMB officials to determine how the improper payment definition specifically applies to the Disaster Assistance Loan Program to determine whether SBA had applied the definition correctly, and interviewed Office of Disaster Assistance (ODA) officials regarding their assessment approach. Our scope and methodology is provided at Appendix I.
We independently validated (shown at Appendix II) SBA’s estimate of improper payments by reviewing a statistical sample of 70 disaster loan disbursements made in FY 2007 to determine whether the disbursements were made in compliance with key SBA approval and disbursement requirements. Based on our review results, we developed an estimate of improper payments in the program. In developing the estimate, we consulted with OMB officials and employed the services of a statistician.
We conducted the audit between August and December 2008 in accordance with Government Auditing Standards prescribed by the Comptroller General of the United States.
The IPIA was enacted to enhance the integrity of the government’s payments and the efficiency of its programs. This legislation, in conjunction with implementing guidance from OMB, requires executive branch agency heads to review their programs and activities annually, identify those that may be susceptible to significant improper payments, estimate amounts improperly paid, and report on the amounts of improper payments and actions to reduce them.
Significant erroneous payments are defined in IPIA as those exceeding both 2.5 percent of program payments and $10 million. An improper payment includes any payment that should not have been made (including a payment to an ineligible recipient), or that was made in an incorrect amount (including overpayments and underpayments) under statutory, contractual, administrative, or legally applicable requirements. Under a direct loan program, improper payments may include disbursements to borrowers or other third-party payments that are based on incomplete, inaccurate, or fraudulent information. They may also include duplicate disbursements, disbursements in the incorrect amount, or loan funds used for purposes other than those allowed by law, program regulations, or agency policy.
In 2003, OMB issued Memorandum 03-13, Implementation Guidance for the Improper Payments Information Act of 2002, which provides agencies withspecific instructions for estimating and reporting the rate of improper payments.
This guidance, which was revised in 2006 when OMB issued Appendix C of Circular A-123, requires agencies to:
Review all programs and activities and identify those susceptible to significant erroneous payments;
Obtain a statistically valid estimate of the annual amount of improper payments in its programs and activities;
Implement a plan to reduce erroneous payments; and
Report estimates of the annual amount of improper payments in its programs and activities and progress in reducing them.
In 2002, OMB identified SBA’s Disaster Loan Program, which provides direct loans to help homeowners, renters, businesses, and non-profit organizations return to pre-disaster condition, as susceptible to improper payments. In FY 2005, SBA first reported an improper payment rate for the program, which it determined was insignificant. Since FY 2005, SBA has reported an improper payment rate for the program ranging from 0.07 to 0.8 percent. Most recently, in the Agency’s 2007 Performance and Activity Report, SBA estimated improper payments to be $4.5 million, or 0.55 percent of the $819.7 million in loans approved in 2007. Several recent OIG audit reports of the Disaster Assistance Loan Program have identified a significant number of improper loan disbursements, causing us to question whether the rate of improper payments reported for the program is accurate. For example:
In May 2007, we reported that SBA issued $368 million in loan disbursements without perfecting liens on property used as collateral, or without completing required UCC filings, as required.(1)
In September 2007, we reported that the Agency approved an estimated $1.5 billion in disaster loans to applicants who lacked repayment ability.(2)
In a June 2008 report on the Agency’s compliance with annual credit review requirements, we reported that SBA disbursed over $1 billion in Gulf Coast Hurricane disaster loan funds at least 1 year after loan approval without verifying whether borrowers had maintained repayment ability and creditworthiness.( 3)
In September 2008, we estimated that SBA approved at least $69 million in loans to borrowers who lacked repayment ability or creditworthiness.(4)
ODA prepares the Agency’s estimate of improper payments for the Disaster Assistance Loan Program. This estimate is based on an annual Quality Assurance Review (QAR) of the program, which includes an improper payment component.
The FY 2007 QAR was performed by seven customer service field representatives, who reviewed a sample of 210 of 18,952 disaster loans that were approved between August 1, 2006, and July 31, 2007.
RESULTS IN BRIEF
SBA’s estimate for FY 2007 significantly understates the level of improper payments in the Disaster Assistance Loan Program and is not statistically valid. SBA reported that improper payments were about $4.5 million, or 0.55 percent of the $819.7 million in loans approved in FY 2007. In contrast, we estimated the improper payment rate to be at least 46 percent, or approximately $1.5 billion of $3.4 billion in loans disbursed in FY 2007. Our estimate, which was developed with the assistance of a statistician and with consult from OMB, demonstrates a rate of improper payments that is so significant it merits an improper payment reduction plan.
We reviewed 70 disbursements and determined that 45 payments were either made to ineligible borrowers or without documentation that is required for disbursement. Of the 45 improper disbursements, 19 were made without complete information, 17 went to ineligible applicants, and 9 involved payments to ineligible applicants without complete information.
ODA, which prepared the estimate for SBA, did not properly estimate the rate of improper payments due to major sampling design errors, flaws in measurement methodology, and inadequate reviews of loan files. Sampling design flaws included using a population of $819.7 million in approved loans, and not the population of $3.4 billion in disbursed loans, as required. This resulted in ODA sampling from a smaller universe. Loan approvals also do not serve as a good sampling population because many loans are cancelled after approval. Finally, ODA employed attribute sampling, when it should have used variable sampling.
ODA also did not properly calculate the error rate in the sample or properly apply that rate in projecting the total value of improper payments that had been made in FY 2007. This occurred because ODA did not consult with a statistician, as required by OMB guidance.
Further, SBA’s improper payment reviews did not adequately detect significant errors in the loan origination and disbursement process. Our review of 30 of the 210 loans sampled by ODA revealed 9 improper payments, where ODA had only identified 2. This occurred because the checklist provided reviewers did not define what constituted an “improper payment” and reviewers did not understand what payments should be categorized as improper.
Because ODA’s FY 2007 improper payment rate was so significantly understated, the Agency needs to develop a corrective action plan to reduce improper payments. Additionally, ODA made similar errors in its FY 2008 improper payment review process.
We recommended that ODA: (1) revise the sampling design methodology for estimating improper payments; (2) consult with a statistician to ensure that the Agency’s sample selection and projection methodology is statistically valid; (3) provide written instructions to QAR reviewers on what constitutes an improper payment; (4) implement a corrective action plan to reduce improper payments; and (5) recalculate the FY 2008 estimate of improper payments, based on audit findings, and if different from the original estimate of improper payments, provide the revised estimate to the Acting Chief Financial Officer.
We also recommended that the Acting Chief Financial Officer: (1) report the improper payment rate calculated by the OIG for FY 2007 to OMB; and (2) ensure the correct FY 2008 improper payment estimate is reported to OMB. Management concurred with the recommendations, but did not provide implementation dates for proposed actions or, in some cases, sufficient detail on planned actions needed to consider its comments responsive. For example, management stated it would consult with OMB to revise its sampling methodology and estimate of improper payments, but did not indicate a time frame for doing so or describe what changes would be made. Management also did not identify additional steps it would take to reduce the high level of improper payments identified by the audit. Therefore, we plan to pursue a management decision on this issue through the audit resolution process.
SBA Significantly Underestimated the FY 2007 Improper Payment Rate ODA significantly underestimated the amount of improper payments in its Disaster Assistance Loan Program. While the FY 2007 improper rate was reported to be 0.55 percent, or approximately $4 million, we estimate it was at least 46 percent, or approximately $1.5 billion.(5) With the assistance of a statistician, we reviewed a statistical sample of 70 disaster loan disbursements made in FY 2007, to determine whether loan recipients met SBA loan eligibility requirements and had submitted all of the required information for loan disbursement. Under the loan program, disbursements are incremental based on the borrowers’ needs and ability to produce evidence of insurance, proper use of prior proceeds, collateral documentation and other information. Loan officers must ensure that the required documents are secured before disbursing the loans.
We determined that 45 of the 70 sampled disbursements were improper. OMB Circular A-123 Appendix C defines an improper payment as including disbursements that were made based on incomplete information as well as disbursements to ineligible recipients. Using the OMB criteria, we determined that 19 disbursements were made without complete information, 17 went to ineligible applicants, and 9 involved payments to ineligible applicants without complete information. A listing of the 45 improper payments is provided in Appendix III.
Disbursements made without complete information lacked either adequate proof of insurance, demonstration of proper use of prior proceeds, adequate proof of insurance injections, or complete collateral documentation. The ineligible applicants either lacked repayment ability or creditworthiness. We discussed these specific deficiencies with OMB officials, who confirmed that they constituted improper payments in a direct loan program(6) according to the guidelines in Appendix C of OMB Circular A-123.
The 19 disbursements that were based on incomplete information lacked evidence of one or more of the following:
Insurance. ODA did not obtain the required evidence of flood and/or hazard insurance. When loans are secured, SOP 50 30 requires hazard insurance for all insurable collateral, including both the damaged and relocation properties. Borrowers must obtain hazard insurance equal to at least 80 percent of the insurable value of the property, and proof of this insurance must be provided to ODA prior to disbursement of loan funds in excess of $10,000. The SOP also requires that borrowers located in a flood zone have flood insurance coverage in the amount equal to the insurable value of the property or the maximum coverage available, whichever is less. Borrowers are required to provide evidence to ODA of the purchase of flood insurance prior to any loan disbursement.
Proper Use of Prior Proceeds. We did not find adequate evidence that previous loan proceeds had been properly spent. SOP 50 30 requires that, prior to any subsequent disbursement where the aggregate amount exceeds $50,000, ODA must obtain evidence that previously disbursed funds were properly used in accordance with the loan authorization agreement. The evidence may be provided in several forms, including joint payee checks, escrow account, progress inspections, lien waivers, or progress certifications.
Insurance Injections. ODA did not obtain sufficient evidence of prior insurance injections. SOP 50 30 requires that all insurance proceeds received, along with any other disaster aid to borrowers, are injected into the repair or replacement of the damaged real estate prior to disbursement of any funds for real estate construction or repair. ODA did not obtain proof that all insurance proceeds received by the borrowers were properly injected prior to disbursement.
Complete Collateral Documentation. ODA made loan disbursements to borrowers without obtaining all required collateral documents. In each case, ODA did not obtain a required mortgage, security instrument, or valid guarantee document. SOP 50 30 allows the Agency to disburse the first $10,000 of secured loans without obtaining collateral documentation.
However, for disbursements up to $50,000, the Agency must at least obtain an executed collateral document, along with a check from the borrower made out to the title recording company. For disbursements over $50,000, ODA must obtain evidence that all required collateral documents have been recorded.
We identified 17 disbursements that were made to ineligible applicants. In two of these cases applicants lacked both repayment ability and creditworthiness.
A summary of these disbursements is provided below:
Repayment Ability. After performing the applicable cash flow computation, we determined that loan disbursements were made to borrowers who lacked the ability to repay their loans. The majority of these loans were approved under an expedited loan approval process, which allowed ODA to approve applicants based on credit score only. However, 13 CFR 123.6 requires SBA to have reasonable assurance that the borrower can repay the disaster loan out of personal or business cash flow. In addition, ODA’s Standard Operating Procedure (SOP) 50 30 requires that cash flow be the basis for establishing repayment ability when approving disaster loans. In a previous audit of the expedited loan approval process, we reported that the process conflicted with Agency regulations by ignoring cash flow. (7)
As a result, the Agency approved $1.5 billion in disaster loans to applicants who lacked repayment ability. The Agency agreed to terminate the program, but never conceded that the program was inappropriate or in conflict with Agency regulations. OMB agreed that disbursements resulting from expedited loans to borrowers who lacked repayment ability constituted improper payments.
Creditworthiness. We identified disbursements made to borrowers with significant credit deficiencies that were not adequately addressed during the loan origination process. 13 CFR 123.6 requires borrowers to have satisfactory credit and character, and SOP 50 30 requires derogatory credit issues to be addressed prior to loan approval. Loan officers are required to obtain a thorough understanding of a borrower’s overall credit history before determining whether credit is satisfactory, and borrowers with poor credit history should be given every opportunity to explain derogatory items before a credit decision is reached. If the loan officer does not explain how the borrower has overcome all significant derogatory credit issues, the loan cannot be approved.
Finally, nine disbursements were made to ineligible applicants without all required documentation. For example, ODA disbursed a loan to one applicant without ensuring that the individual had repayment ability, had used insurance proceeds to repair damaged property, properly applied prior loan proceeds, and had proof of hazard insurance.
ODA Used an Invalid Sample Design and Measurement Methodology and Did Not Adequately Review Sampled Loans
OMB Circular A-123, Appendix C requires agencies to obtain a statistically valid estimate of the annual amount of improper payments for programs that are most susceptible to erroneous payments. Agencies must also consult with a statistician to ensure the validity of their sample design, sample size, and measurement methodology. In some cases, agencies may need to use more complex sample designs because their universe contains differing dollar amounts.
The primary reasons that ODA inaccurately estimated its FY 2007 improper payments were because it:
Used an invalid sampling design, which incorrectly defined the population and was inappropriate for the population sampled;
Employed a statistically invalid measurement methodology to estimate the rate of improper payments; and
Relied on a loan review process that did not adequately detect errors constituting improper payments.
SBA did not consult with a statistician at any point during its improper payment review process. Had the Agency done so, it may have used a valid sampling design and measurement methodology. The loan review process was inadequate because the individuals conducting the reviews lacked an understanding of what constituted improper payments.
ODA Used an Invalid Sampling Design
SBA incorrectly used a universe of approved loans instead of disbursements to obtain its statistical sample of the FY 2007 disaster loan activity, contrary to OMB guidance. OMB Circular A-123, Appendix C states that under a direct loan program, improper payments include disbursements to borrowers or other third party payments that are based on incomplete, inaccurate, or fraudulent information. According to OMB, agencies may use loan approvals as the universe if it is representative of, or comparable to, the amount of disbursements made that year. In this case, approvals were not representative of annual disbursements because in FY 2007 SBA approved $819.7 million in loans, but disbursed $3.4 billion. The $3.4 billion included disbursements made on loans approved in FY 2007 as well as disbursements on loans approved in prior years. Another reason why approvals were not an appropriate universe from which to sample was because $175.5 million, or 21 percent, of the approved loans were cancelled, and therefore, never disbursed. Consequently, the total dollar value of loans approved did not correspond to the actual amount disbursed by the Agency.
Further, disaster loans generally are not fully disbursed upon approval. Agency policy that was in effect during FY 2007 allowed recipients to receive $10,000 at the time of loan closing with subsequent disbursements based on borrower needs and evidence of insurance, proper use of prior proceeds, and other information.
For this reason, some improper payments may only be identified when determining whether subsequent disbursements are appropriate. By sampling from a universe of only approved loans, ODA effectively eliminated from consideration improper payments that were associated with subsequent disbursements. For example, we identified disbursements that were made on loans approved prior to FY 2007 that should not have been made because ODA had not secured proof of insurance, collateral documents, and other information required for the disbursement. Finally, using approvals as the sampling universe is inappropriate because many loans are cancelled prior to approval. As previously discussed, the agency cancelled 21 percent of the loans approved in FY 2007.
Additionally, ODA used a methodology that was inappropriate for the population it was testing. Specifically, ODA employed attribute sampling, which is a technique that is used to determine whether a characteristic exists in the population. For example, this methodology could be used to test whether a payment was improper, but it is inappropriate for estimating the rate or value of improper payments. Variable sampling is generally used to predict a value for a given population. This technique involves sampling a number of payments computing the value of those payments, and finally deriving the statistical projection of the value of those payments in the population. Although attribute sampling is not designed for estimating the rate or value of the population, ODA used it to derive its estimate of improper payments.
ODA Employed a Statistically Invalid Measurement Methodology
We also determined that ODA did not properly calculate the error rate in the sample that was used to project the rate of improper payments. One of the components used in the calculation was the total value of loans in the sample, which included loans that were cancelled subsequent to approval. As a result, ODA used a value of $14.9 million when it should have been $13.3 million. This had the effect of generating a smaller error rate.
Further, in making its projections ODA combined data from different time periods, using one time period for its sample error rate and another for the population that the rate was applied against. Specifically, the error rate was obtained from a sample of loans approved between August 1, 2006 and July 31, 2007, but was projected to a universe of loans approved between October 1, 2006 and September 30, 2007. Finally, ODA added a $2,564 improper payment that was outside of the random sample in making its projection. Consequently, its projection methodology was statistically invalid.
ODA Relied on a Loan Review Process that Did Not Adequately Detect Errors
A review of 30 of the 210 loans sampled by ODA disclosed that ODA had not identified all of the improper payments associated with these loans. ODA identified only 2 improper payments in the 30 loans we sampled, while we identified a total of 9 (including the 2 identified by ODA), or 30 percent of the sampled loans.
The additional loans we identified as being improper payments included loans that went to ineligible borrowers and/or payments that were made without securing the documentation required for disbursement. For example, one improper payment went to an individual who was ineligible because he had repayment and credit issues and had already been partially reimbursed for the damage by FEMA. Therefore, this payment was also a duplicate benefit. Another payment went to an individual that was ineligible based on her poor credit history. ODA also did not timely secure a recorded mortgage for this damaged property. ODA’s review of the loans was conducted as part of the annual QAR process. As part of this process reviewers complete a checklist, which includes a question about whether the loan processing decision resulted in an erroneous payment.
However, the checklist does not define what constitutes an “improper payment” to ensure that reviewers consider all eligibility issues as well as data completeness.
Consequently, reviewers were completing the checklist without an understanding of what payments should be categorized as improper. We interviewed five of the seven individuals who performed the FY 2007 review. None of them were familiar with the Improper Payments Information Act, knew the definition of an improper payment according to OMB guidelines, or were aware that the results of their work would be used to estimate the Agency’s improper payment rate. When asked to define an improper payment, the reviewers provided varying answers. Three of the five believed an improper payment occurred when the approved monthly loan payment was incorrect. The reviewers also indicated no written guidance had been provided to them on how to conduct the improper payment component of the QAR.
Such guidance would have been particularly helpful in determining repayment ability as the checklist does not instruct reviewers to recalculate borrower repayment ability when portions of an individual’s income and/or debt cannot be confirmed.
Agency management provided us with a summary of the FY 2008 improper payment review process. Although we did not perform a full analysis of the FY 2008 estimate, we determined that the Agency again sampled loan approvals instead of disbursements, and used the same error rate calculation and projection methodology.
Because ODA did not employ a statistically valid methodology or sample from the correct population, its estimate of improper payments is significantly understated, and a revised estimate should be reported to OMB. The improper rate calculated by the OIG shows that the level of improper payments in the Disaster Assistance Loan Program is significant, thus meriting a corrective action plan. Further, because ODA employed the same methodology in computing its improper payment estimate for FY 2008, it should recalculate the FY 2008 estimate and report the revised rate to OMB, if the estimate differs from the original calculation.
We recommend that the Chief, Executive Office of Disaster Strategic Planning and Operations direct the Associate Administrator for Disaster Assistance to:
Revise the sampling design methodology for estimating improper payments to ensure that the sample is based on disbursements versus loan approvals in accordance with OMB guidance, and employs variable sampling procedures.
Require that a statistician be consulted when developing the sample error rate and projection methodology to ensure that the estimate derived is statistically valid, as required by OMB guidance.
Provide written instructions to QAR reviewers on what constitutes an improper payment, as defined by OMB Circular A-123, Appendix C, and which directs reviewers to recalculate borrower repayment ability when income and/or debt cannot be confirmed, or as appropriate.
Implement a corrective action plan to reduce improper payments in the Disaster Assistance Loan Program.
Recalculate the FY 2008 estimate of improper payments, based on audit findings, and if different from the original estimate of improper payments, provide the revised estimate to the Acting Chief Financial Officer.
We also recommend that the Acting Chief Financial Officer:
Report the improper rate calculated by the OIG for FY 2007 to OMB; and
Ensure the correct FY 2008 improper payment estimate is reported to OMB.
AGENCY COMMENTS AND OFFICE OF INSPECTOR GENERAL RESPONSE
On February 5, 2009, we provided SBA with a draft of the report for comment. On March 6, 2009, ODA submitted its formal response, along with endorsement by the Chief of the Executive Office of Disaster Strategic Planning and Operations, which is contained in its entirety in Appendix IV. On March 17, 2009, OFA submitted its formal response, which is also contained in its entirety in Appendix IV. ODA generally concurred with the recommendations, but commented on several issues raised in the report. The CFO’s office concurred with our recommendations. Management’s comments are summarized below, with ODA’s specific comments summarized by the headings in its response.
Comment 1 - Sample Design and Measurement Methodology
Management stated that our recommendation that improper payment reviews be based on loan disbursements instead of loan approvals is reasonable, but is not absolutely required by OMB guidance. While the OIG consulted with OMB on this matter, management stated that it has not received official instructions from OMB directing it to sample from loan disbursements instead of loan approvals.
Further, management believes the OMB guidance is unclear and not absolute. Management stated that it will revisit the issue with OMB and obtain clear guidance on how the improper payment definition applies to the Disaster Loan Program.
We believe that ODA already has official guidance from OMB on how its sample should be pulled. OMB Circular A-123, Appendix C states “that under a direct loan program, improper payments may include disbursements to borrowers or other third parties that are based on incomplete, inaccurate, or fraudulent information.” OMB guidance also allows agencies to use loan approvals as the universe if it is representative, or comparable to, the amount of disbursements made on loans approved in the fiscal year as well as on loans made in prior years. Therefore, while we agree that the OMB guidance is not absolute, we believe it is clear. In this case, loan approvals was not an appropriate universe from which to sample as it was not comparable to the amount of FY 2007 disbursements made on new and previously approved loans. In FY 2007 SBA approved $819.7 million in loans, but disbursed $3.4 billion on new and prior year loans. We will provide any requested guidance to ODA management; we simply followed OMB guidance for our audit, as discussed in our exit briefing. To ensure that improper payments are accurately determined, we agree with ODA’s decision to revisit this issue and obtain clear guidance from OMB. Based on the OMB guidance, ODA should inform our office of its findings and any revised improper payment rates reported for fiscal years 2007 and 2008.
Comment 2 - Loans Processed Under the Expedited Approval Process
Management stated that we should not have characterized loans disbursed under the previous expedited process as improper payments since they were made in compliance with Agency guidelines. Based on guidance in place at that time, the program authorized loans to applicants based solely on satisfactory credit ratings, even though individuals may have lacked repayment ability.
We disagree that improper payments associated with the Expedited Disaster Loan Program should not be considered, simply because ODA guidance established for the pilot improperly permitted loans to be issued without a repayment analysis. ODA clearly implemented a policy for the Expedited Disaster Loan pilot program that was contrary to SBA regulations. 13 CFR § 123.6 requires that in approving disaster loans, there must be reasonable assurance that borrowers can repay their loans out of their personal or business cash flow. Therefore, for ODA to insist that payments made without establishing applicant repayment ability should be considered proper contradicts Agency regulations on applicant eligibility. As defined by OMB, loans made to ineligible individuals constitute improper payments. In recognition that the Expedited Program was issuing loans that should not have been approved, SBA terminated the program and replaced it with one that required an analysis of applicant repayment ability.
Comment 3 – Disbursements Lacking Proper Documentation
Management claims that the OIG cited deficiencies on payments made without receipts and proof of insurance, even though these payments were proper.
We question how ODA was able to determine whether payments were properly made if, as it acknowledges, actions were not properly documented in the loan files. During the comment period, ODA had the opportunity to, but did not provide information to support its position that such payment actions were proper. Given the significant differences between improper payments reported by the Agency and that identified by the audit, we believe SBA’s estimate must be corrected and resubmitted to OMB.
SBA will consult with OMB to obtain clear official guidance before revising the sampling design methodology for estimating improper payments.
We believe that pursuing additional guidance from OMB will help the Agency in revising its sampling methodology for its Disaster Loan Program. However, because management’s response did not provide specific steps it will take to implement the recommendation or a target date for implementation, we do not consider it to be fully responsive.
SBA will comply with OMB Circular A-123 and consult with a statistician.
Management’s comments were not fully responsive to the recommendation because a target date for implementing the proposed action was not provided.
QAR reviewers will be provided with written instructions that will include a definition of improper payments and guidance on documenting repayment ability.
The action proposed by management will address the recommendation, but because management’s response did not provide a target date for completion of this action, we do not consider it to be fully responsive.
Management stated that it has already implemented new training, procedures, and aggressive quality assurance reviews, and will continue to document steps taken to eliminate or reduce the exceptions identified in this audit.
Management’s comments were not responsive to the recommendation as no additional actions were proposed to reduce the significant rate of improper payments in the Disaster Loan Program. Management’s statement that improperly documented disbursements and those made under expedited procedures to individuals who lacked repayment are not improper payments infers that the Agency plans no further actions on those loans. We remained concerned that if no action is taken, additional improper payments may be made on those loans, which have not yet been fully disbursed. For this reason, we will pursue resolution of the recommendation through the audit resolution process, and will provide a copy of our report to OMB.
SBA will revise the FY 2008 estimate of improper payments after consultation with OMB to obtain official guidance regarding the sampling design methodology and criteria for estimating improper payments for the Disaster Loan Program.
Management’s comments are not fully responsive to our recommendation as a target completion date was not provided. We also request that ODA inform our office of its findings, any program revisions, and the time frame for revising the FY 2008 estimate.
OIG RECOMMENDATIONS AND THE CFO’S RESPONSE
Management stated that after additional guidance is obtained from OMB, it will issue an appropriate disclosure on previously reported improper payment rates.
Management proposed an alternative action to that recommended. Instead of reporting the improper payment rate estimated by the OIG, it plans to consult first with OMB in determining what disclosures should be made. Because it is uncertain what disclosures SBA plans to make, we do not consider management’s comments to be responsive at this time, and request that a management decision on this recommendation be provided after it has consulted with OMB.
SBA will report its FY 2009 improper payment rate to OMB in November 2009 using the MAX reporting system. At that time, SBA will include an appropriate disclosure related to the previously reported FY 2008 improper payment rate.
CFO’s comments are responsive to our recommendation. However, we request that the CFO’s office provide us with a description of the “MAX reporting system” and its findings regarding the revised FY 2008 improper payment estimate for the Disaster Loan Program.
We request that ODA submit written comments identifying (1) the specific actions it will take to implement recommendations 1 and 5, and (2) the target dates for completion of actions in response to recommendations 1, 2, 3, and 5. We will pursue resolution of recommendation 4 through the audit resolution process. We request that the CFO’s office submit its (1) findings regarding both recommendations 6 and 7, (2) specific actions planned to implement them, (3) target dates for completion of these actions, and (4) notification of the FY 2009 rate to OMB in November 2009. We would appreciate receiving your additional comments within fifteen days of the final report date.
We appreciate the courtesies and cooperation of the Chief, Executive Office of Disaster Strategic Planning and Operations; Office of Associate Administrator, Disaster Assistance; Office of the Chief Financial Officer; Disaster Assistance Processing and Disbursement Center and DCMS Operations Center representatives during this audit. If you have any questions concerning this report, please call me at (202) 205- [FOIA ex. 2] or Pamela Steele-Nelson, Director, Disaster Assistance Group, at (202) 205- [FOIA ex. 2].
To view the appendix items, please refer to the document attached below.
1 OIG Report 07-22, Securing Collateral For Disaster Loan Disbursements, May 9, 2007.
2 OIG Report No. 07-34, The Quality of Loans Processed Under the Expedited Disaster Loan Program,
September 28, 2007.
3 OIG Report 08-10, Annual Credit Reviews For Gulf Coast Hurricane Disaster Loan Disbursements,
June 17, 2008.
4 OIG Report 08-19, Early-Defaulted Gulf Coast Hurricane Disaster Loans, September 12, 2008.
5 This estimate is based on a statistical projection at a 95 percent confidence level.
6 OMB provided its opinion in question/answer format, in response to a set of specific questions.
7 OIG Report No. 07-34, Audit of the Quality of Loans Processed Under the Expedited Disaster Loan Program,
September 28, 2007.