2b = Law Enforcement Guideline(s)
2d = Law Enforcement Technique(s)
2e = Law Enforcement Procedure(s)

Executive Summary

Objective and Scope

Background

Results

Conclusion

Appendix I – Detailed Objective, Scope, and Methodology

Appendix II – Major Contributors to This Report

Appendix III – Report Distribution List

Appendix IV – Outcome Measures

Executive Summary

The General Accounting Office and the Treasury Inspector General for Tax Administration have repeatedly reported on the high number of fraudulent refunds claimed on tax returns. In October 1994, the Department of the Treasury’s Under Secretary (Enforcement) testified before the Congress that up to "$5 billion in problematic refunds [was being] paid out annually." The Internal Revenue Service (IRS) established the Questionable Refund Program in 1977 to detect and stop fraudulent claims for refunds on income tax returns. The program’s major operations occur in the service centers by Criminal Investigation’s (CI) Questionable Refund Detection Teams (QRDT).

Our review assessed the effectiveness of selected activities of the QRDTs. Specifically, we determined whether the QRDTs effectively and consistently used available information to identify fraudulent returns and refund schemes. Further, we determined whether fraudulent returns were properly controlled and fraudulent refunds were stopped, as appropriate.

CI management can strengthen controls to better ensure fraudulent returns are identified and refunds are stopped, as required. Actions can be taken to improve the effectiveness of QRDT activities. Specifically:

Controls should ensure that fraudulent tax refunds ****2b,2e****. Our review of 669 fraudulent returns identified by QRDT personnel in calendar year 1998 showed that 50 of these returns involved refunds where the taxpayer had a ****2b,2e****. In 35 (70 percent) of these cases, QRDT personnel did not ****2b,2e****.

The Electronic Fraud Detection System (EFDS) is the major fraud detection tool for electronically filed returns with a separate database located at each of the five service centers that process electronically filed (e-file) returns (Andover, Memphis, Austin, Cincinnati, and Ogden). The Treasury Enforcement Communications System (TECS) is a nationwide system available to QRDT personnel to initially identify schemes perpetrated in more than one service center (i.e., multi-center schemes). QRDT personnel generally researched their local EFDS database to identify fraudulent returns. The four remaining EFDS databases were not always accessed to identify fraudulent returns. Additionally, QRDTs did not always use the TECS.

Generally, QRDT personnel properly controlled fraudulent returns by entering the required information (i.e., transaction codes) on the IRS computer system. However, these controls were not always removed from the IRS computer system as required. Our sample of 400 cases showed that computer controls for 20 cases (5 percent) should have been removed from the IRS computer system. Further, QRDT personnel did not always apply consistent criteria when determining the specific computer controls used to monitor taxpayers who have filed fraudulent returns, and the length of time these computer controls are maintained on taxpayer accounts.

Guidelines specify that QRDT personnel should verify the validity of a refund within 10 working days of delaying its issuance to a taxpayer. We obtained information from the IRS computer system showing instances where QRDT personnel, at the four sites visited, delayed the issuance of suspicious refunds during the period February through December 1998. Our review of 100 of these cases showed that 43 refunds were not verified and issued to taxpayers within 10 working days as specified. These refunds were issued to taxpayers from 11 to 166 working days after they were delayed, with an average of 42 days.

CI management can improve QRDT activities by establishing national guidelines detailing the process staff should follow to ****2b,2e****. Guidelines should also be developed to ensure that personnel use established controls designed to monitor the activity of taxpayers who previously filed fraudulent returns in a consistent manner. QRDT personnel should better use information available on the EFDS and the TECS to identify fraudulent refund schemes. Finally, CI management should re-emphasize the need to verify suspicious returns as fraudulent or legitimate in a timely manner, and release refunds to taxpayers when appropriate.

Management Response: Criminal Investigation management agreed with the facts contained in a discussion draft report previously provided. However, the IRS’ official response was not available for inclusion at the time this final report was issued. We provided the IRS management with a draft of this report on November 1, 1999, with a 30 calendar-day comment period.

Objective and Scope

 We performed this audit to provide management with an assessment of the Internal Revenue Service’s (IRS) efforts to identify fraudulent returns and to stop fraudulent refunds as part of the Revenue Protection Strategy.

Our overall objective was to assess the effectiveness of selected activities of the Questionable Refund Detection Teams (QRDT). Specifically, we assessed whether the QRDTs effectively and consistently used available information to identify fraudulent returns and refund schemes. Additionally, we determined whether fraudulent returns were properly controlled and fraudulent refunds were stopped as required.

To accomplish this objective, we:

• Interviewed QRDT personnel at four service centers to determine the process followed to identify fraudulent returns, prevent the issuance of fraudulent refunds, and identify other fraudulent returns that are related to a fraud scheme.
• Analyzed fraudulent returns identified by QRDT at four service centers in 1998 to determine whether the returns were properly controlled and the refunds were stopped as required.
• Assessed whether QRDT personnel properly and timely released refunds for questionable returns subsequently determined to be non-fraudulent and reversed case controls as appropriate.

Our audit work was performed at the National Office Criminal Investigation (CI) function and four service centers (Andover, Philadelphia, Ogden, and Fresno). This review was performed in accordance with Government Auditing Standards. Audit work was conducted from December 1998 through June 1999.

Details of our audit objective, scope, and methodology are presented in Appendix I. Major contributors to this report are listed in Appendix II.

Background

The Questionable Refund Program (QRP) is a nationwide multi-functional program established in January 1977. The purpose of this program is to detect fraudulent returns and stop the payment of false refunds claimed on income tax returns. From January 1977 through December 1998, CI statistics show that QRP has been responsible for the detection of over $962 million in fraudulent refunds and has been successful in stopping 78 percent of these refunds.

The program’s major operations occur in the service centers. The QRDTs, under the supervision of the Chief, Criminal Investigation Branch (CIB), identify questionable refund returns and other non-compliance issues. Particular emphasis is placed on preventing the issuance of refund checks and direct deposits to multiple and fraudulent filers. Specifically, the QRDTs conduct preliminary reviews of questionable returns identified by manual and computerized screening techniques. Those returns with substantive indications of being fraudulent are referred to field offices for criminal investigation. Returns not meeting criminal investigation criteria are referred to other IRS functions for civil actions, when appropriate.

Results

CI management can strengthen controls to better ensure fraudulent returns are identified and refunds are stopped, as appropriate. The effectiveness of QRDT activities can be improved by:

• Ensuring that fraudulent tax refunds ****2b,2e****.
• Using all available information to identify nationwide fraud schemes.
• Administering computer controls over fraudulent returns in a more consistent and effective manner.
• Ensuring that refunds for suspicious returns are verified as fraudulent or legitimate in a more timely manner and issued to taxpayers as required.

A QRP training manual was provided to QRDT personnel containing guidelines related to various aspects of QRP operations, including identifying fraudulent returns, stopping fraudulent refunds, and identifying fraud schemes. Although this document is in draft status, National Office CI management stated that QRDT personnel should be following the procedures outlined in this document.

Controls Can Be Improved to Better Ensure that Fraudulent Refunds ****2b,2e****

 CI management can better ensure that fraudulent refunds ****2b,2e****. Refunds determined to be fraudulent by QRDT personnel were ****2b,2e****.

****2b,2e****. The IRS inputs specific transactions (Transaction Codes (TC) 916/918) to its computer system on all taxpayer accounts that have been identified in connection with the filing (or attempted filing) of a fictitious return. At the time of our review, ****2b,2e****. Subsequently, QRDT personnel will identify fraudulent returns and input TC 916/918 controls on taxpayer accounts as required.

QRDT personnel identified 24,564 fraudulent returns during processing year 1998 at the 4 service centers where audit work was performed. We reviewed a random sample of 669 of these fraudulent returns to determine whether the cases were controlled as appropriate and the refund was stopped. Refunds associated with 50 of these returns were ****2b,2e****. In 35 (70 percent) of these cases, QRDT personnel ****2b,2e****. These offsets totaled over $59,000, or an average of approximately $1,700 per return. CIB personnel claimed over $50,000 of these ****2b,2e**** as a program accomplishment on management reports (i.e., refund not issued to the taxpayer).

To further assess the potential scope of this issue, we obtained a computer download to identify the number of returns input to the IRS computer system in calendar year 1998 that were under TC 916/918 control and ****2b,2e****. We identified 3,387 returns that met this criteria. In 2,559 (76 percent) of these cases, ****2b,2e****. These refunds totaled over $3.1 million.

There are no national guidelines ****2b,2e****. QRDT personnel at one service center generally ****2b,2e****. QRDT personnel at another location historically ****2b,2e****. Local procedures at two service centers require that ****2b,2e****. The national analyst responsible for QRDT operations stated that ****2b,2e****.

The use of fraudulent refunds ****2b,2e**** could cause embarrassment to the IRS and diminishes its efforts to deter fraud. Program accomplishments could be overstated if ****2b,2e****.

1. Establish national guidelines detailing the process QRDT personnel should follow ****2b,2e****.

2. Review the cases identified by the Treasury Inspector General for Tax Administration on IRS computer records to determine whether the ****2b,2e****.

Management Response: Criminal Investigation management agreed with the facts contained in a discussion draft report previously provided. However, the IRS’ official response was not available for inclusion at the time this final report was issued. We provided the IRS management with a draft of this report on November 1, 1999, with a 30 calendar-day comment period.

Information Available to the Questionable Refund Detection Teams Could Be Better Used to Identify Fraudulent Returns

 CI management can better ensure that fraudulent return schemes involving returns filed at more than one service center are identified. The QRDTs, at the service centers visited, generally do not research national tax return information maintained on the Electronic Fraud Detection System (EFDS) when attempting to identify fraudulent tax returns. In addition, the QRDTs do not always use the Treasury Enforcement Communications System (TECS) as specified.

The EFDS is the major fraud detection tool for electronically filed returns. A separate EFDS database is located at each of the five service centers that process electronically filed (e-file) returns (Andover, Memphis, Austin, Cincinnati, and Ogden). Most QRDT personnel only have access to the EFDS database at the service center handling e-file returns for their location.

The query feature of the EFDS enables QRDT personnel to search for information meeting specific criteria to identify fraudulent returns. Return information for the current filing season is specific to each EFDS database. Specifically, each EFDS database contains tax return information related to returns filed electronically and certain returns filed on paper at that service center, as well as certain returns filed on paper at selected service centers within its jurisdiction.

We determined that QRDT personnel generally research only their local EFDS database to identify fraudulent returns. The four remaining EFDS host databases generally are not accessed to identify fraudulent returns. As a result, QRDT personnel at each service center generally do not research information related to returns filed at either 7 or 8 of the 10 service centers when attempting to identify related fraudulent returns.

We selected a random sample of 10 schemes identified during the 1999 filing season at 2 service centers to determine whether related returns could be identified through research of the 4 other EFDS databases. Our research identified one fraudulent return filed at another service center that was related to one of these schemes. QRDT management indicated that this return was not identified because they had not obtained password access to this EFDS database.

A limited number of QRDT personnel at each CIB have access to all EFDS databases. The responsible national EFDS analyst believed that it is possible to enable every tax examiner to have access to all EFDS databases when performing scheme development. However, technical issues such as adequate memory capacity and system response time would need to be addressed.

The TECS is a nationwide system available to QRDT personnel to initially identify schemes perpetrated in more than one service center (i.e., multi-center schemes). The QRDTs use the TECS for transmitting scheme information to other service centers and storing historical information in the System. However, our work showed that information related to fraudulent returns is not always researched on the TECS as specified.

The QRP training manual specifies that QRDT personnel research the TECS for certain return information for all fraudulent returns identified. In addition, information from one return for each fraud scheme should be entered into the TECS.

QRDT personnel at two of the four service centers visited do not research the TECS as specified. QRDT personnel at one of these sites research the TECS for only one return from each scheme instead of all returns. QRDT management at the other site stated that the TECS has not been used at all during this filing season due to a lack of resources.

When all available information is not used, fraudulent returns may go undetected. More frequent access to national EFDS and TECS information may better ensure that fraudulent returns are identified and refunds are stopped as required.

3. Ensure that QRDT personnel who are currently authorized to access the five EFDS databases research this information on a more frequent basis when identifying fraudulent returns.
4. Assess the technical modifications necessary to enable additional QRDT personnel to access national EFDS information as appropriate.
5. Ensure that the TECS is used, as appropriate, to better identify fraudulent returns.

Management Response: Criminal Investigation management agreed with the facts contained in a discussion draft report previously provided. However, the IRS’ official response was not available for inclusion at the time this final report was issued.

Computer Controls Over Fraudulent Returns Can Be More Effectively and Consistently Administered

 Generally, QRDT personnel properly controlled fraudulent returns by entering the required information (i.e., transaction codes) on the IRS computer system. However, these controls were not always removed from the IRS computer system, as required. Further, QRDT personnel did not always apply consistent criteria when determining the specific computer controls used to monitor these taxpayers and the length of time these computer controls were maintained on taxpayer accounts.

The IRS inputs specific transactions to its computer system on all taxpayer accounts that have been identified in connection with the filing (or attempted filing) of a fraudulent return. These transactions, known as TC 916 and TC 918, are used to ensure that all transactions related to the tax year under control are systemically identified by the IRS and reviewed by a QRDT. TC 916 suspends all activity on a taxpayer’s account for a specific tax year and prevents a refund from being issued. TC 918 is similar to a TC 916, but suspends activity on a taxpayer’s account for all tax years. TC 916/918 should be input to the IRS computer system for all questionable refund investigations.

We reviewed a random sample of 669 fraudulent returns identified by the QRDTs during calendar year 1998 at the 4 service centers visited to determine whether case controls were input as required. Forty-five cases did not require TC 916/918 control because they were classified as Return Preparer Program (RPP) schemes. Per national guidelines, QRDT personnel are not required to input IRS computer controls on RPP cases. Of the remaining 624 cases, QRDT personnel controlled 621 with either a TC 916 or TC 918.

To further assess whether potentially fraudulent returns were properly controlled, we obtained information from the IRS computer system showing instances where QRDT prevented the issuance of suspicious refunds. During the period February through December 1998, QRDT personnel, at the 4 sites visited, delayed the issuance of 9,963 suspicious refunds. We determined refunds associated with 1,165 of these cases were not issued to the taxpayer but had no indications of a TC 916/918 control. QRDT personnel verify the income information contained on suspicious returns by ****2b,2d****. If the ****2b,2d**** prior to the issuance of the refund, QRDT personnel must determine whether to stop the refund before the issuance date. TC 916/918 controls should not be used for suspicious returns.

Our sample of 100 of the 1,165 cases showed that 8 were not processed as appropriate. Specifically, three cases should have been controlled by a TC 916/918. The refund for three additional cases should have been released to the taxpayer. QRDT personnel stated that further research was necessary on two cases to determine if the cases should be controlled or whether the refund should be released to the taxpayer.

QRDT personnel could more consistently use these controls. Specifically, the QRDTs at two service centers generally use TCs 918 to control taxpayer accounts associated with the filing of fraudulent returns. The QRDTs, at the other two service centers visited, use TCs 918 only in limited situations.

The responsible national analyst stated the choice of a TC 916 or a TC 918 is a local decision. National guidance issued to the field in November 1994 recommends the use of TC 918 on ****2b,2d****.

The QRDT receives the Control Case Inventory Report each quarter containing all tax periods under
TC 916/918 control. These reports should be reconciled quarterly by QRDT personnel, and used to validate that all controls are on the taxpayer account or removed as appropriate.

CI management has provided limited guidance on the appropriate duration to maintain TC 916/918 controls. In December 1992, the time limit for the retention of TC 916/918 controls was eliminated from national guidelines. The intent was to allow QRDTs to maintain controls on accounts for longer, or shorter, than the previous restriction of two subsequent years from the year the fraudulent return was detected.

QRDT personnel at two service centers reverse TC 918 controls ****2b,2d****. QRDT personnel at another service center reverse a TC 918 control ****2b,2d****. QRDT personnel at another site do not follow a standardized process when determining whether case controls should be reversed.

We obtained the most recent quarterly report available at the time of our visitation containing all cases under
TC 918 control at the four service centers. We selected a judgmental sample of 100 cases under TC 918 control at each site (400 total cases) to determine whether the control should be removed from the taxpayer’s account. We determined that TC 918 controls should have been reversed in 20 of 400 cases analyzed. We used the ****2b,2d**** as a basis for the removal of the TC 918.

QRDT personnel agreed that these controls should have been reversed. QRDT management at one service center indicated that the quarterly report is reconciled if time permits. Although QRDT management at the other service centers stated that these reports are reviewed quarterly, we could not verify this activity because documentation of their review is not maintained in CIB files or reviewed by management. QRDT personnel at each site stated that they did not receive any guidance on how to perform this verification.

Inconsistent taxpayer treatment could result when the choice of a TC 916/918 control and the timing of their reversal is left to the discretion of individual QRDTs. When a TC 918 is unnecessarily maintained on a taxpayer’s account, the QRDT workload increases as all tax returns filed by the taxpayer must be reviewed for accuracy. Further, this control could increase taxpayer burden as legitimate refunds could be unnecessarily delayed pending QRDT review.

6. Establish national procedures related to the use of TC 916/918 to ensure consistent treatment of taxpayers.
7. Develop procedures to provide clear guidelines to CIBs on how the quarterly Control Case Inventory Reports should be reconciled, and when case controls should be removed.

Management Response: Criminal Investigation management agreed with the facts contained in a discussion draft report previously provided. However, the IRS’ official response was not available for inclusion at the time this final report was issued.

Refunds for Suspicious Returns Can Be More Timely Verified as Fraudulent or Legitimate and Released by Questionable Refund Detection Team Personnel as Required

Refunds associated with suspicious returns were not always verified as fraudulent or legitimate within specified time frames and issued to taxpayers as appropriate. QRDT personnel did not always ensure suspicious refunds were verified within 10 working days and issued to taxpayers, as appropriate, for 43 of 100 cases analyzed.

QRDT personnel verify the income information contained on suspicious returns by ****2b,2d****. The QRP training manual states that this verification should be performed by telephone. Faxes can also be used but require a quick response time from ****2b,2d****. Although the manual specifies that the use of letters to verify income should be kept to a minimum, the National Office CI analyst responsible stated that the QRDTs were advised not to use any letters to ****2b,2d**** for suspicious returns.

If the ****2b,2d**** prior to the issuance of the refund, QRDT personnel must determine whether to stop the refund before the issuance date. If the refund has been stopped, the QRP training manual specifies that QRDT personnel should verify these returns within 10 working days and release the refund for returns determined to be legitimate. Additionally, the National Office CI analyst responsible stated that refunds should also be released if QRDT personnel are not able to verify the income within 10 days to protect the rights of innocent taxpayers.

We obtained information from the IRS computer system showing instances where CIB delayed the issuance of suspicious refunds. During the period February through December 1998, CIB personnel, at the 4 sites visited, delayed the issuance of 9,963 suspicious refunds. We determined refunds associated with 2,881 of these cases were subsequently released to the taxpayer after QRDT personnel delayed their issuance.

Our sample of 100 of these cases showed that 43 refunds were not released to taxpayers within 10 working days. These refunds were released to taxpayers from 11 to 166 working days after they were delayed, with an average of 42 days.

QRDT personnel at two sites visited were unaware that the QRP training manual specifies that suspicious returns should be verified as false or valid within 10 working days and refunds should be released on valid returns. QRDT personnel at one site sent over 2,000 letters ****2b,2d**** to verify income for suspicious returns. If a suspicious return cannot be verified within one week, local procedures at another site require QRDT personnel to ensure the refund is stopped and consider the return fraudulent.

When refunds associated with suspicious returns are not timely verified as fraudulent or legitimate and issued as appropriate, IRS efforts to provide quality customer service to taxpayers is adversely affected.

8. Re-emphasize to QRDT personnel the need to verify refunds for suspicious returns within established guidelines and release legitimate refunds as appropriate.

Management Response: Criminal Investigation management agreed with the facts contained in a discussion draft report previously provided. However, the IRS’ official response was not available for inclusion at the time this final report was issued.

Conclusion

 CI management can strengthen controls to better ensure fraudulent returns are identified and refunds are stopped as appropriate. The effectiveness of QRDT activities can be improved by ensuring that fraudulent tax refunds ****2b,2d****. Additionally, QRDTs should use all available information to identify nationwide fraud schemes, and administer computer controls over fraudulent returns in a more consistent and effective manner. Finally, refunds for suspicious returns should be verified as fraudulent or legitimate in a more timely manner and issued to taxpayers as appropriate.

Appendix I

The objective of the review was to assess the effectiveness of selected activities of the Questionable Refund Detection Teams (QRDT). We assessed whether the QRDT is effectively and consistently using available information to identify fraudulent returns and refund schemes. Additionally, we determined whether fraudulent returns are properly controlled and refunds are properly deleted.

To accomplish the overall objective, we performed the following audit tests:

1. Determined whether the Internal Revenue Service effectively identifies cases with substantial indicators of fraud in a consistent manner.
1. Determined if QRDT operations at the four service centers visited were effectively and consistently utilizing applicable research tools to identify refund schemes and prevent the issuance of fraudulent refunds.
2. Interviewed management personnel at selected service centers to determine methods used to obtain and review return information, identify fraudulent returns, stop the issuance of fraudulent refunds, develop fraud schemes, and refer these cases to the field for investigation.
3. Obtained and analyzed national procedures to identify any differences between the actual process followed at each service center and methods specified by the National Office.
4. Determined whether the scheme development process can be improved to better use QRDT resources and ensure all fraudulent returns related to known schemes are properly identified.

1. Assessed whether the QRDT activities ensure that returns related to a scheme are properly identified.
1. Determined whether QRDT performs queries on all Electronic Fraud Detection System (EFDS) databases to identify returns filed in other service centers related to known schemes.
2. Interviewed the Treasury Enforcement Communications System Coordinator at each service center to assess how this system is used to identify fraudulent returns.

3. Selected a random sample of 10 of 53 fraud schemes identified during the 1999 filing season by 2 of the 4 service centers visited to determine whether additional related returns could be identified through research of applicable EFDS databases.

1. Determined if identified fraudulent returns are properly controlled and refunds are stopped in a timely manner, when appropriate.

1. Reviewed a random sample of 669 of 24,564 fraudulent returns identified in processing year 1998 at the four service centers visited to determine whether the returns were properly controlled and the refunds were stopped.
2. Determined the local policy for utilizing specific computer controls designed to monitor taxpayers that previously filed fraudulent returns.
3. Assessed whether QRDT personnel properly and timely released refunds for questionable returns subsequently determined to be non-fraudulent and reversed case controls as appropriate.
0. Randomly sampled 100 of 1,165 cases without a computer control where QRDT personnel at the four service centers visited stopped a refund to determine whether a control should be input. The QRDT delayed the issuance of refunds for these cases during the period February through December 1998.
1. Randomly sampled 100 of 2,881 cases where QRDT personnel at the four service centers visited stopped a refund but later released it to determine whether the release was timely. The QRDT delayed the issuance of refunds for these cases during the period February through December 1998.
2. Judgmentally sampled 400 of 22,391 cases from the four selected service centers’ most recent quarterly report showing all cases under Transaction Code 918 control to determine whether the control should have been removed from the taxpayer’s account. The most recent quarterly report available at the time of our review was the fourth quarter for calendar year 1998 (three service centers) and first quarter for calendar year 1999 (one service center).

Appendix II

Walter Arrison, Associate Inspector General for Audit (Wage and Investment Income Programs)

Kerry Kilpatrick, Director

Mark Nathan, Audit Manager

Cheryl Cerqua, Senior Auditor

Jeffrey Jones, Senior Auditor

Larry Madsen, Senior Auditor

John O’Rourke, Senior Auditor

Margaret Anketell, Auditor

Theresa Berube, Auditor

Roberta Bruno, Auditor

Annette Hodson, Auditor

Michael McGovern, Auditor

William Richards, Auditor

Larry Wyrick, Auditor

Appendix III

Deputy Commissioner Operations C:DO

Chief Operations Officer OP

Assistant Commissioner (Criminal Investigation) OP:CI

Executive Officer for Service Center Operations OP:SC

National Director for Legislative Affairs CL:LA

Director, Andover Submission Processing Center D

Director, Fresno Submission Processing Center D

Director, Ogden Submission Processing Center D

Director, Philadelphia Submission Processing Center D

Office of the Chief Counsel CC

Office of Management Controls M:CFO:A:M

Director, Office of Program Evaluation and Risk Analysis M:OP

Appendix IV

This appendix provides information on the measurable impact that our recommended corrective actions will have on tax administration. These benefits will be incorporated into our Semiannual Report to the Congress.

Finding and recommendations:

Controls can be improved to better ensure that fraudulent refunds ****2b**** (page 3). Recommendations were made to:

• Establish national guidelines detailing the process staff should follow ****2b****.
• Review the cases identified by the Treasury Inspector General for Tax Administration on Internal Revenue Service (IRS) computer records to determine whether the ****2b****.

Type of Outcome Measure:

Revenue Protection - (potential)

Value of the Benefit/Methodology:

QRDT personnel identified 24,564 fraudulent returns during calendar year 1998 at the 4 service centers where audit work was performed. We reviewed a random sample of 669 of these fraudulent returns to determine whether the cases were controlled as appropriate and the refund was stopped. Refunds associated with 50 of these returns were ****2b****. In 35 (70 percent) of these cases, QRDT personnel ****2b****. These offsets totaled over $59,000, or an average of approximately $1,700 per return. Criminal Investigation Branch (CIB) personnel claimed over $50,000 of these ****2b**** as a program accomplishment on management reports (i.e., refund not issued to the taxpayer).

To further assess the potential scope of this issue, we obtained a computer download to identify the number of returns input to the IRS computer system in calendar year 1998 that were under Transaction Code (TC) 916/918 control and ****2b****. We identified 3,387 returns that met this criteria. In 2,559 (76 percent) of these cases, ****2b****. These refunds totaled over $3.1 million.

Findings and recommendations:

Computer controls over fraudulent returns can be more effectively and consistently administered (page 8). Additionally, refunds for suspicious returns can be more timely verified as fraudulent or legitimate and released by QRDT personnel as appropriate (page 12). Recommendations were made to:

• Establish national procedures related to the use of TC 916/918 to ensure consistent treatment of taxpayers.
• Develop procedures to provide clear guidelines to CIBs on how the quarterly Control Case Inventory Report should be reconciled and when case controls should be removed.
• Re-emphasize to QRDT personnel the need to verify suspicious returns as fraudulent or legitimate in a timely manner and release refunds to taxpayers when appropriate.

Type of Outcome Measure:

Taxpayer Entitlements – (actual)

Value of the Benefit/Methodology:

We selected a judgmental sample of 100 cases under TC 918 control at each site (400 total cases) from a universe of 22,391 cases to determine whether the control should be removed from the taxpayer’s account. We determined that TC 918 controls should have been reversed in 20 of 400 cases analyzed.

We obtained information from the IRS computer system showing instances where CIB delayed the issuance of suspicious refunds. During the period February through December 1998, CIB personnel at the four sites visited delayed the issuance of 9,963 suspicious refunds. We determined refunds associated with 2,881 of these cases were subsequently released to the taxpayer after QRDT personnel delayed their issuance.

Our random sample of 100 of these 2,881 cases showed that 43 refunds were not released to taxpayers within 10 working days. These refunds were released to taxpayers from 11 to 166 working days after they were delayed, with an average of 42 days.