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

Appendix V – Other Issues Identified During Review

Appendix VI – Overview of Pipeline Processing

Appendix VII – Key Processes

Appendix VIII – Management’s Initial Response to the Draft Report

Appendix IX – Management’s Amended Response to the Draft Report

Executive Summary

The Internal Revenue Service (IRS) is the largest processor of data in the world. The majority of the data for individual tax returns is processed during the first half of each calendar year, commonly referred to as a "filing season," when individual taxpayers file their returns. For the 2000 Filing Season, the IRS faced considerable challenges, including legislated tax law changes; changes in IRS staffing; and updates to IRS computers and equipment, including Year 2000 (Y2K) compliance.

The objective of this audit was to evaluate the effectiveness of the IRS’ procedures for processing paper tax returns filed by individual (non-business) taxpayers. This review concentrated on evaluating the adequacy of the IRS’ process for tracking filing season activities and reacting to problems as they occurred. In a separate review, the Treasury Inspector General for Tax Administration (TIGTA) will provide an assessment of the IRS’ processing of electronically filed individual tax returns.

The IRS effectively processed individual paper tax returns filed during the 2000 Filing Season. As of April 28, 2000, the IRS had processed over 54 million paper individual tax returns. Although there were some isolated problems, the IRS’ 2000 Filing Season was a success.

The IRS had not programmed its computer system to properly process changes related to the Child Tax Credit (CTC) and the Additional Child Tax Credit (ACTC). In addition, the IRS modified its computer programming designed to verify Secondary Social Security Numbers (S-SSN). As a result, the IRS cannot ensure that taxpayers are compliant with these tax laws. Also, the database the IRS used to track tax law programming changes was not being updated regularly, thus it was not providing assurance that required tax law programming changes were completed.

The IRS did not identify potentially unqualified CTC claims. This credit, established for tax years beginning in 1998, allows taxpayers to reduce their tax liabilities by $400 ($500 in 1999) for each qualifying child. A qualifying child must be age 16 or below and meet other requirements, such as being a citizen or resident of the United States and claimed as a dependent on the taxpayer’s return.

We identified over 750,000 1998 and 1999 tax returns with over $339 million of potentially unqualified CTCs that could not be supported by IRS date of birth information. These potentially unqualified credits were allowed because the IRS had postponed computer programming changes designed to validate the date of birth for CTC claims every year since the CTC was authorized (1998). The IRS advised us that this programming was postponed due to higher priorities (e.g., Y2K programming) and limited programming resources.

We identified and reported this issue to the IRS in March 2000. At that time, the amount of the potentially unqualified claims for 1999 was $10 million on 23,000 returns. However, IRS officials advised us that they were not in a position to implement any corrective actions at that time because of the process used to implement programming changes to its computer systems. By April 6, 2000, the amount of the potentially unqualified credits for 1999 had increased to $54 million on 101,000 returns.

The postponed computer programming changes designed to validate the date of birth for the CTC also prevented the IRS from identifying potentially erroneous claims for the ACTC. The ACTC is allowed for taxpayers who claim three or more children and meet additional qualifying requirements, such as limitations in other credits on the return (e.g., adoption credit, mortgage interest credit, etc.). The ACTC may result in a refund to the taxpayer even if no tax is due.

We identified 33,000 tax returns that had over $12 million in potentially unqualified ACTC claims. By not programming its computers to validate date of birth information, the IRS allowed ACTCs for some dependents who were over the acceptable age.

We presented the CTC and ACTC information to IRS management, who acknowledged that their current tools for tracking filing season activities were not able to verify this information. They requested specific examples to determine what collection actions, if any, should be taken. Due to legal limitations, the IRS may not be able to collect the potentially erroneous refunds resulting from the unqualified ACTC claims. However, collection actions can be initiated on the potentially erroneous refunds resulting from the unqualified CTC claims.

In 1996, the Congress passed legislation requiring the IRS to ensure that taxpayers provided correct social security numbers on their tax returns. S-SSNs are generally the social security numbers for the spouse on a joint tax return. According to the IRS’ own estimates, approximately 2.7 to 3 million taxpayers have invalid S-SSNs.

The IRS initially planned to validate all S-SSNs processed during the 2000 Filing Season. However, based on early volumes of returns received with invalid S-SSNs, the IRS determined that continued processing of these returns would overload the computer systems it uses to resolve errors. Since these returns could be processed with invalid S-SSNs, the IRS approved computer programming changes to bypass the errors due to invalid S-SSNs when the associated SSNs on the return are valid. The IRS also developed a plan to send notices to these taxpayers after the 2000 Filing Season. However, these actions did not ensure that the IRS implemented the legislative requirement.

A prior TIGTA audit report on the 1999 Filing Season recommended the IRS conduct regular updates to the database used to track tax law programming changes as a means to document completed actions. We determined that the database was not being updated regularly, thus it was not providing assurance that required tax law programming changes were completed.

The IRS should implement computer programming changes to ensure that taxpayers comply with tax laws relating to the CTC, the ACTC, and S-SSNs and initiate actions to recover the potential lost revenue due to the postponed programming for these issues. The IRS should ensure that its computer system has the capacity to correctly process tax returns and should send notices to taxpayers with invalid S-SSNs in sufficient time to allow for corrections. Finally, the IRS should ensure the database used to track tax law programming changes is updated regularly.

Management’s Response: The IRS agreed with three of our five recommendations. Of the remaining two recommendations, the IRS disagreed with one recommendation and questioned the legality and cost benefit of implementing the other recommendation. The IRS believes that the current capacity of the computer system used to store errors found on returns filed by taxpayers is sufficient and, therefore, does not plan to take any corrective action. The IRS stated that additional legal guidance would be needed regarding the legality of attempting to recover the CTC mentioned in the report and also stated that the collection costs would likely exceed the amount of overpayments.

Management’s complete responses to the draft report are included in Appendices VIII and IX.

Office of Audit Comment: Several discussions we had with IRS executives indicated they had concerns with the computer system used to store errors found on returns and its inability to handle the expected fallout of errors from verifying the S-SSNs during the 2000 Filing Season. However, since we did not test the computer system’s capacity during this review, we will address this issue during our audit of the 2001 Filing Season.

Also, we encourage the IRS to focus on recovering the refunds resulting from the unqualified CTC claims. Unlike the ACTC cases, collection assessments may be made on these cases.

The IRS also initially disagreed with the $351 million outcome measure concerning the potential lost revenue represented by unqualified CTC and ACTC claims. As stated in the report, we could not electronically verify the ages of dependents when more than four were claimed on a tax return; these returns must be manually reviewed. This was due to the fact that the IRS only maintains four date of birth fields on the Returns Transaction File (RTF). Because of this, we revised our outcome measure and will only claim $211 million for the CTC claims with four or less dependents. The IRS agreed with this revised outcome measure.

Objective and Scope

This review was conducted as part of the Treasury Inspector General for Tax Administration (TIGTA) Fiscal Year 2000 audit plan. The overall objective of the review was to evaluate the effectiveness of the Internal Revenue Service’s (IRS) procedures for processing paper tax returns filed by individual (non-business) taxpayers. The review concentrated on evaluating the adequacy of the IRS’ process for tracking filing season activities and reacting to problems as they occur. In a separate review, the TIGTA will provide an assessment of the IRS’ processing of electronically filed individual tax returns.

The audit was conducted from January to April 2000 at the National Headquarters and the Andover, Atlanta, Austin, Fresno, and Kansas City Submission Processing Centers. Submission Processing Centers process tax returns and payments received from taxpayers. The audit was performed in accordance with Government Auditing Standards.

We selected and reviewed judgmental samples of returns and payments received at the above Submission Processing Centers to determine if they were processed correctly. We also held discussions with various IRS personnel at the above centers and the National Headquarters to determine how they tracked and resolved problems identified during the filing season. 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 IRS is the largest processor of data in the world. Most of the data for individual tax returns is processed during each "filing season." In general, the IRS defines the filing season as the first half of each calendar year, when individual taxpayers file their tax returns. The IRS receives and processes returns through a nationwide network of 10 Submission Processing Centers. Generally, paper tax returns and related correspondence are received at the Submission Processing Centers, checked for errors, and input to the taxpayers’ accounts on the IRS’ computer system. Payments are deposited into the Federal Reserve Bank as quickly as possible. If a taxpayer has paid more tax than they owe, the IRS issues a refund to the taxpayer. If a taxpayer has not paid all the tax due, the IRS sends the taxpayer a notice requesting payment for the balance due. The IRS also sends the taxpayer a notice if an error was made on the return. The notice generally explains why the error occurred and any resulting balance due or refund owed.

For the 2000 Filing Season, the IRS faced considerable challenges, including legislated tax law changes; changes in IRS staffing; and updates to IRS computers and equipment, including Year 2000 (Y2K) compliance. Legislative changes require the IRS to re-program its computers to check for taxpayer errors. One of these legislative changes, the IRS Restructuring and Reform Act of 1998 (RRA 98), allowed taxpayers to claim a $400 child tax credit (in 1998) for each qualifying child age 16 and below. The credit was effective for tax years beginning after December 31, 1997. For tax year 1999, the credit was increased to $500 per qualifying child.

The Personal Responsibility and Work Opportunity Reconciliation Act and the Small Business Job Protection Act require taxpayers to provide valid social security numbers (SSNs) for themselves, their spouses, and their dependents. The spouse’s SSN is generally referred to as the Secondary SSN on a tax return. If taxpayers do not provide valid SSNs, the IRS can disallow certain items on their tax returns. These acts were passed to deny certain credits and benefits to individuals who were not authorized to claim them. This provision applies to returns due after September 21, 1996.

Results

As of April 28, 2000, the IRS had processed over 54 million paper individual tax returns. Although there were some isolated problems, the IRS’ 2000 Filing Season was a success (see Appendix V for issues reported to management concerning these isolated problems and management’s responses).

Within these successes, we identified opportunities for the IRS to more effectively implement tax law changes. The IRS had not programmed its computers to properly process changes related to the Child Tax Credit (CTC) and the Additional Child Tax Credit (ACTC). In addition, the IRS modified its computer programming designed to verify Secondary Social Security Numbers (S-SSN). Also, the database used to track tax law programming changes was not being updated regularly, thus it was not providing assurance that required programming changes were completed.

Implementation of improvements described in this report could have prevented the IRS from allowing over $351 million of potentially unqualified CTC and ACTC to taxpayers. These improvements would also help ensure that taxpayers comply with the law.

The Internal Revenue Service Had a Successful 2000 Filing Season

Despite the many challenges it faced, the IRS had a very successful 2000 Filing Season. We reviewed some of the key processes used by the IRS to track returns and payments from the time they were received in the Submission Processing Centers until the information was input to the taxpayers’ accounts on the IRS’ computer system. Our review of a judgmental sample of over 1,700 paper individual tax returns and over 300 related payments received at the 5 Submission Processing Centers early in the 2000 Filing Season showed the following:

• Most taxpayer errors were corrected properly.
• Payments received with returns were deposited timely and posted properly to taxpayers’ accounts.
• Tax returns were posted properly to taxpayers’ accounts.

See Appendix VII for details on the key processes reviewed to ensure that returns were effectively processed.

Opportunities Exist for the Internal Revenue Service to More Effectively Implement Tax Law Changes

We determined that the IRS could do a better job implementing three tax law changes that were in effect during the 2000 Filing Season. These laws related to the:

• Child Tax Credit (CTC).
• Additional Child Tax Credit (ACTC).
• Secondary Social Security Numbers (S-SSN).

Details for each of these areas follow.

The IRS had not programmed its computers to validate CTC claims. The CTC was established by law and was effective for tax years beginning in 1998. The law allows that, for each qualifying child, taxpayers can take a credit of $400 ($500 in 1999) directly against the amount of tax owed. If no tax is owed, then no CTC is allowed. A qualifying child must be under the age of 17, a citizen or resident of the United States, and claimed as a dependent on the taxpayer’s return, either as his/her child, stepchild, adopted child, grandchild, or eligible foster child.

We identified over 750,000 tax returns with over $339 million of potentially unqualified CTCs that could not be supported by IRS date of birth information. These results include Tax Year 1998 and 1999 returns processed from January 1, 1999, through April 6, 2000.

By not programming its computers to validate these claims, the IRS allowed CTCs for some dependents who were over the acceptable age. The chart below shows the age breakdown of the dependents for our 1999 CTC analysis.

Although most of the 750,000 tax returns were for taxpayers claiming more CTC than they were qualified for, over 90,000 of these returns were for taxpayers claiming less CTC than they were qualified for. For example, some cases showed zero CTC claimed; however, the IRS changed the zero and allowed the CTC based on the number of dependents listed on the return. The IRS did not check the date of birth for the dependents on the 90,000 returns to ensure they qualified for the CTC.

The CTC was allowed in both of the scenarios mentioned above because the IRS had postponed computer programming designed to validate the date of birth for CTC claims every year since the CTC was authorized. The IRS advised us that programming was postponed due to higher priorities and limited programming resources.

We identified and reported this issue to IRS management in March 2000. At that time, the amount of the potentially unqualified CTC for 1999 was $10 million on 23,000 returns. However, IRS officials advised us that they were not in a position to implement any corrective actions at that time because of the process used to implement programming changes. The process used requires that requests for computer programming changes be submitted almost a year in advance. By April 6, 2000, the amount of the potentially unqualified CTCs for 1999 was $54 million on 101,000 returns.

The postponed computer programming changes designed to validate the date of birth for the CTC also prevented the IRS from identifying potentially erroneous claims for the ACTC. The ACTC is allowed for taxpayers who claim three or more children and meet additional qualifying requirements, such as limitations in other credits (e.g., adoption credit, mortgage interest credit, etc.). The ACTC may result in a refund to the taxpayer even if no tax is due.

Among the 750,000 tax returns with potentially unqualified CTC claims, we identified 33,000 returns with over $12 million in ACTC claims that could not be supported by the information contained in IRS date of birth information (i.e., some of the dependents on these returns were over the acceptable age for claiming the ACTC). The following hypothetical example illustrates this condition:

Due to legal limitations, the IRS may not be able to initiate actions to collect the potentially erroneous refunds resulting from the unqualified ACTC claims. Because the ACTC is a refundable credit versus a non-refundable credit (e.g., the CTC), different collection procedures exist. However, collection actions can be taken on the potentially erroneous refunds resulting from the unqualified CTC claims.

We presented the CTC and ACTC information to IRS management, who acknowledged that their current tools used to track filing season activities were not able to verify or dispute this information. They requested specific examples to determine what collection actions, if any, should be taken.

Other CTC and ACTC issues were reported in prior TIGTA audit reports and memoranda dating back to 1999. The prior issues related to (1) the clarity of notices sent to taxpayers explaining errors made when claiming the CTC, and (2) the need to post-review cases affected by specific legislative provisions and to conduct trend analyses of the results to identify needed changes or improvements to tax forms and instructions. The CTC continues to be a problem area for the IRS.

During the 2000 Filing Season, the IRS modified its computer programming designed to verify S-SSNs on tax returns. This action was taken due to storage limitations on the computer system it uses to correct errors on tax returns. According to the IRS’ own estimates, approximately 2.7 to 3 million taxpayers have invalid S-SSNs. The S-SSN is generally the SSN for the spouse on a joint tax return.

In 1996, the Congress passed legislation requiring the IRS to validate name controls to individual SSNs for taxpayers on their tax returns. Originally, the IRS planned to validate all S-SSNs processed during the 2000 Filing Season. These plans included sending notices to taxpayers with invalid S-SSNs in December 1999. The notices were to inform taxpayers of the need to ensure that all SSNs (primary, secondary, and dependent) reported on their returns were correct. This action would help reduce the possibility of returns being rejected during processing because of incorrect SSNs. IRS management expected these notices to reduce the number of returns filed with invalid S-SSNs.

However, due to the anticipated volume of potential Social Security Administration (SSA) contacts, the SSA requested that the IRS delay issuance of these notices. In response, IRS management agreed not to send the notices until after the 2000 Filing Season but went ahead with its plans to identify and correct invalid S-SSNs on returns processed during the 2000 Filing Season.

During the first few weeks of the 2000 Filing Season, the Submission Processing Centers were faced with storage limitation problems on the computer system they use to correct errors. This problem occurred because of the number of returns being received with invalid S-SSNs. Since these returns could be processed with invalid S-SSNs, IRS personnel approved a computer programming change to bypass the errors due to invalid S-SSNs when the associated SSNs were valid. For the returns with invalid S-SSNs that were not identified during the 2000 Filing Season, IRS management developed a detailed plan to identify the invalid S-SSNs during the summer and send taxpayers notices in October 2000.

Although this action allowed the IRS to minimize the impact of these returns on its computer system, it did not ensure that the IRS implemented the legislative requirement.

The IRS needs to resolve the capacity limitations of the computer system it uses to correct errors. If computer programming for the CTC is completed for the 2001 Filing Season, it is doubtful that the IRS computer system used to resolve errors on returns can accommodate the volume of invalid S-SSNs and erroneous CTC and ACTC claims.

Without adopting a proactive approach to identify and prioritize returns with errors, the IRS will continue to face the risk of having to react to unexpected challenges due to the storage limitations of the computer system it uses to correct errors.

A prior TIGTA audit report on the 1999 Filing Season recommended that the IRS conduct regular updates to the database used to track tax law programming changes as a means to document completed actions. We determined that the database was not being updated regularly, thus it was not providing assurance that required tax law programming changes were completed.

For example, the first record of the database lists programming for the CTC as the required action. The approval status column shows that IRS personnel agreed to program for the CTC. The database shows the due date for programming the CTC as February 28, 1998, and the last update as September 23, 1999. The actual date for postponing the CTC programming change was May 24, 1999, almost 4 months before the last CTC update to the database. The database had not been updated because of other higher priority work.

Without continued monitoring and updates, the database designed to track legislated programming changes is ineffective.

The Assistant Commissioner (Forms and Submission Processing) should work with other appropriate IRS officials to ensure implementation of the following recommendations:

1. Implement computer programming changes needed to ensure taxpayers comply with the tax laws relating to the CTC, the ACTC, and S-SSNs.

Management’s Response: The IRS responded that, "RIS TSF-0-0016A01 was resubmitted for the 2001 Filing Season and scheduled for implementation January 2, 2001. This RIS will perform a systemic check of the dependent DOB (Date of Birth) on all returns claiming the CTC. The programming to verify the S-SSNs was in place for the 2000 Filing Season. However, due to concerns from the SSA and other internal considerations, management decided to bypass the programming for the 2000 Filing Season. The programming will resume for the 2001 Filing Season."

1. Initiate appropriate actions designed to provide assurance that the IRS computer system is able to store all errors on returns filed by taxpayers.

Management’s Response: The IRS responded that, "We do not agree with this recommendation. We designed the ERS database to hold records for correction and Quality Assurance. The system counts the number of records being loaded. Once the count reaches 109,925, an overflow file will begin loading. This failsafe process prevents the ERS Database Load Program from aborting before reaching the maximum record capacity of 110,000 records per day. We believe this system solves the problem and no corrective action is needed."

Office of Audit Comments: This recommendation was made based on several discussions with IRS executives who mentioned concerns with the ERS and its inability to handle the expected fallout of errors from verifying the S-SSNs during the 2000 Filing Season. However, since we did not conduct testing of the ERS capacity during this review, we will address this issue during our audit of the 2001 Filing Season.

1. Initiate appropriate collection actions to recover the potential lost revenue due to the postponed programming of the CTC date of birth validity checks.

Management’s Response: The IRS responded that, "The IRS Assistant Chief Counsel (Income Tax & Accounting) issued a memorandum in September 1999 concluding the IRS may not recover, through assessment, an erroneous ACTC refund. Further guidance from Chief Counsel would be necessary regarding the legality of attempting to recover the CTCs as recommended in this report. We also question the wisdom of pursuing recovery of these payments because the costs of collection are likely to exceed the amount of these overpayments."

Office of Audit Comments: We reviewed the memorandum from the Assistant Chief Counsel and agree that the IRS may not recover an erroneous ACTC refund through assessment. However, the memorandum does state that, as in the case of any other nonrefundable credit, an overstatement of the nonrefundable portion of the CTC results in an assessable deficiency. Therefore, we believe the IRS can pursue collection actions to recover the lost revenue related to the nonrefundable portion of unqualified CTC claims.

1. Send notices to taxpayers with invalid S-SSNs in sufficient time to allow taxpayers to make corrections.

Management’s Response: The IRS responded that, "We developed and issued an information notice (or soft notice) to taxpayers that filed tax returns with invalid S-SSNs. The IRS began issuing the notices September 5, 2000, which will give SSA enough time to work with taxpayers so they can correct their SSNs before the 2001 Filing Season."

1. Ensure the database used to track tax law programming changes is updated regularly.

Management’s Response: The IRS responded that, "The spreadsheet will no longer be used. Instead, we will use the Legislative Implementation Tracking System (LITS), which was used to track the RRA 98 legislation changes. The LITS is maintained in the Office of Tax Administration Coordination. The system is web based, efficient, and functional. All organizations and functions within the IRS will have access to it, unlike the spreadsheet used previously. The LITS has automated features and is designed to require periodic updates, which will satisfy the concerns raised in the current audit report."

Conclusion

The IRS had a very successful 2000 Filing Season. While this is a noteworthy accomplishment, opportunities exist for the IRS to more effectively implement tax law changes affecting the filing seasons. The IRS needs to ensure that its computer system is able to identify items on returns that are not allowable by law. Specifically, improvements are needed to identify erroneous CTC and ACTC claims and to validate the accuracy of S-SSNs.

Detailed Objective, Scope, and Methodology

The overall objective of the review was to evaluate the effectiveness of the Internal Revenue Service’s (IRS) procedures for processing paper tax returns filed by individual (non-business) taxpayers. We conducted the following tests to achieve this objective:

1. Defined and evaluated the adequacy of the process used to track production results of paper returns filed early during the 2000 Filing season.
1. Interviewed various National Headquarters personnel to identify concerns and management techniques for monitoring the 2000 Filing Season.
2. Interviewed local personnel at the Andover, Atlanta, Austin, Fresno, and Kansas City Submission Processing Centers to identify concerns and management techniques for keeping track of filing season activities.
3. Created a flowchart of the key controls used to keep track of the 2000 Filing Season’s early production results.
2. Assessed the effectiveness of IRS actions to prevent and control paper tax return processing problems.
1. Attended the 2000 Filing Season Production meetings at the Atlanta, Andover, Austin, and Kansas City Submission Processing Centers and created a log of issues.
2. Reviewed correspondence regarding the Service Center Automated Mail Processing System (SCAMPS) and the dual-label initiative.
3. Discussed the SCAMPS and the dual-label initiative with various National Headquarters personnel.
4. Reviewed results of the 2000 Filing Season readiness process and created a log of significant unresolved production problems for the 2000 Filing Season.
5. Prepared a matrix to identify 2000 Filing Season problems to determine if the IRS properly prioritized, communicated, and resolved the issues.
3. Assessed the adequacy of IRS actions to identify and resolve paper tax return processing problems.
1. For each of the 4 weeks from January 25, 2000, through February 18, 2000, reviewed a judgmental sample of a minimum of 100 paper tax returns and remittances from the following functions at the 5 Submission Processing Centers shown in step I.B.
1. Receipt and Control Branch
1. Selected a judgmental sample of 549 tax returns from the Receipt and Control branches at 5 Submission Processing Centers.
2. Reviewed, researched, and manually re-computed information on the returns to determine if amounts were properly calculated.
3. Traced the returns through the document control system (pipeline) to determine if errors/problems were identified and corrected.
4. Used various tools to research, trace, and monitor the processing of paper tax returns and payments.
5. Selected a judgmental sample of 55 of the sampled returns from step 1.a) above and determined whether they were properly maintained in storage facilities after being processed through the pipeline.
2. Code and Edit Section
1. Performed cursory reviews of 4,123 returns (1040 series) to identify returns that had entries affected by recent tax law changes.
2. Identified and reviewed 529 returns from step 2.a) above and conducted research to determine if they posted correctly (i.e., errors were properly corrected and refunds were issued timely).
3. Error Correction Section
1. Selected a judgmental sample of 169 returns (1040 series) from the Error Resolution System Unit.
2. Conducted research to determine the types of errors on the return, if the error was corrected properly, if the correct notice was issued to the taxpayer, and if the taxpayer received a timely refund.
3. Discussed discrepancies with appropriate Submission Processing Center personnel.
4. Remittance Processing Section
1. Selected a judgmental sample of 462 returns (1040 series) containing checks from the Receipt and Control Branch that were received between January 25, 2000, and February 17, 2000.
2. Obtained copies of the returns and the related checks.
3. Conducted research to determine the dates the payments posted to the taxpayers’ accounts.
4. Conducted research to determine whether the payments and returns posted to the taxpayers’ accounts.
5. Reviewed discrepancies between IRS records and the taxpayers’ returns and analyzed the error codes to determine whether discrepancies were adequately resolved.
2. Created a summary of the status of the 2000 Filing Season using various management reports.
1. Sent questionnaires to the Management Information System for Top Level Executives coordinators at the Atlanta, Austin, Andover, Kansas City, and Fresno Submission Processing Centers to determine procedures for validating data forwarded to the National Headquarters.
2. Held discussions with National Headquarters personnel to determine the process used to track year 2000 Filing Season activities for paper returns and the process used to prioritize, control, and resolve any production problems identified.
3. Analyzed weekly data extracts of live tax return transactions for the 10 weeks from January 31, 2000, through April 6, 2000, to determine if tax law changes were processed correctly.
1. Secondary Social Security Number (S-SSN) Validation
1. Researched past legislative changes concerning the S-SSN validation to determine actions required for IRS implementation.
2. Reviewed past and current requests for computer programming changes regarding the S-SSN validation to identify proposed changes.
3. Held discussions with National Headquarters personnel to determine required implementation action by the IRS.
2. Child Tax Credit (CTC) and Additional Child Tax Credit (ACTC)
1. Researched IRS publications, work manuals, and past legislative changes concerning the CTC and ACTC to determine implementation actions to be taken by the IRS and the criteria to qualify for the credits.
2. Obtained data extracts for all of 1999 and the 10-week period ending April 6, 2000, of tax return transactions in which the amount of CTC allowed could not be supported by IRS date for birth information.
3. Analyzed data extracts obtained in step b) to identify both the number and the amount of potentially unqualified CTCs and ACTCs allowed by the IRS.
4. Evaluated the adequacy of IRS corrective actions regarding recent or unresolved Treasury Inspector General for Tax Administration, General Accounting Office, and Federal Managers’ Financial Integrity Act findings regarding the processing of paper tax returns.
1. Determined the current status of prior findings related to contingency plans and related contract modifications with commercial banks.
2. Determined if SCAMPS-related problems identified during the recent readiness review were resolved.
1. Interviewed National Headquarters personnel to determine actions taken to address SCAMPS concerns from the Submission Processing Centers.
2. Reviewed National Headquarters correspondence regarding the SCAMPS and the dual-label initiative.
3. For a 10-workday period from February 11, 2000, to February 25, 2000, selected a judgmental sample of 4,064 return envelopes sorted by the SCAMPS as containing a check to determine if the SCAMPS was accurately detecting checks in envelopes.
4. For an 11-workday period from February 11, 2000, to February 28, 2000, selected a judgmental sample of 1,850 returns that used the non-refund mail labels to determine if the SCAMPS was accurately detecting checks in envelopes.

Major Contributors to This Report

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

M. Susan Boehmer, Director

Deann Baiza, Audit Manager

Don Butler, Audit Manager

Richard Calderon, Audit Manager

Deborah H. Glover, Audit Manager

Gary L. Young, Audit Manager

Deborah Carter, Senior Auditor

Gregory A. Dix, Senior Auditor

Kenneth C. Forbes, Senior Auditor

Carola A. Gaylord, Senior Auditor

Edith R. Lemire, Senior Auditor

Robert A. Olson, Senior Auditor

Carl H. Parmer, Senior Auditor

Tina M. Parmer, Senior Auditor

Anthony W. Anneski, Jr., Auditor

Robert A. Baker, Auditor

Arnita F. Brown, Auditor

Karen C. Fulte, Auditor

Robert J. Howes, Auditor

Glory Jampetero, Auditor

Mary L. Keyes, Auditor

George E. Millard, Auditor

John P. Ojeda, Auditor

Sharla J. Robinson, Auditor

Joseph C. Butler, Computer Specialist

Report Distribution List

Deputy Commissioner N:DC

Commissioner, Wage and Investment Operating Division W

Director, Submission Processing W:CAS:SP

Director, Legislative Affairs CL:LA

National Taxpayer Advocate TA

Office of Management Controls N:CFO:F:M

Office of the Chief Counsel CC

Director, Office of Program Evaluation and Risk Analysis N:ADC:R:O

Director, Submission Processing, Andover W:CAS:SP:AN

Director, Submission Processing, Atlanta W:CAS:SP:AT

Director, Submission Processing, Austin W:CAS:SP:AU

Director, Submission Processing, Fresno W:CAS:SP:F

Director, Submission Processing, Kansas City W:CAS:SP:KC

Director, Strategy and Finance W:S

Outcome Measures

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

Finding and recommendation:

The Internal Revenue Service (IRS) had not programmed its computers to properly process changes related to the Child Tax Credit (CTC) and the Additional Child Tax Credit (ACTC). In addition, the IRS modified its computer programming designed to verify Secondary Social Security Numbers (S-SSNs) (see page 5). We recommended that:

• The IRS implement computer-programming changes needed to ensure taxpayers comply with the tax laws relating to the CTC, the ACTC, and S-SSNs.
• The IRS initiate actions designed to provide assurance that its computer system is able to store all errors found on returns filed by taxpayers.
• The IRS initiate appropriate collection actions to recover the potential lost revenue due to the postponed programming of the CTC date of birth validity checks.
• The IRS send out notices to taxpayers with invalid S-SSNs in sufficient time to allow taxpayers to make corrections.

Child Tax Credit Claims - The IRS had not programmed its computers to identify potentially erroneous CTC claims. We identified over 750,000 tax returns in which the IRS allowed taxpayers over $339 million of potentially unqualified CTCs. These results include Tax Year 1998 and 1999 returns processed from January 1, 1999, through April 6, 2000. The results affected approximately 710,000 taxpayers since some taxpayers received unqualified CTCs on both 1998 and 1999 tax returns.

These CTCs were allowed because the IRS had postponed computer programming designed to validate the date of birth for CTC claims since the CTC was authorized.

Additional Child Tax Credit Claims - The postponed computer programming changes designed to validate the date of birth for the CTC also prevented the IRS from identifying potentially erroneous claims for the ACTC. Among the 750,000 tax returns with potentially unqualified CTC claims, we identified 33,000 tax returns (affecting approximately 26,000 taxpayers) with over $12 million in ACTC claims that could not be supported by the information contained in IRS date of birth information.

Secondary Social Security Numbers - During the 2000 Filing Season, the IRS modified its computer programming, designed to verify S-SSNs on tax returns. This action was taken due to storage limitations on the computer system the IRS uses to correct errors on tax returns. However, this action did not ensure the IRS implemented the legislative requirement to verify S-SSNs on tax returns. According to the IRS’ own estimates, approximately 2.7 to 3 million taxpayers have invalid S-SSNs.

Type of Outcome Measure:

Increased Revenue/Revenue Protection - Potential

Value of the Benefit:

By not programming its computers to validate CTC claims, the IRS did not protect over $339 million of revenue due to potentially unqualified CTCs allowed to approximately 710,000 taxpayers during the period January 1, 1999, to April 6, 2000. However, because we could not electronically verify the ages of dependents when more than four were claimed on a tax return, we are only claiming $211 million in outcomes for 505,706 taxpayers. (See methodology section for further details.)

Methodology Used to Measure the Reported Benefit:

Our analysis is based on information obtained from computer queries of the IRS Returns Transaction File (RTF) for Tax Years 1998 and 1999. The methodology for our computer query/analysis is as follows:

The query first identified taxpayers who were allowed the CTC on their 1998 and 1999 returns by selecting records having an amount greater than zero in the ‘Child Tax Credit Amount Computer’ field of the RTF. The query then counted, for each taxpayer receiving the CTC, the number of dependents with date of birth values within the range to be qualified for the CTC (i.e., legitimate dependents). For the 1998 analysis, the range was 1998 to 1982; for 1999, the range was 1999 to 1983 (i.e., under the age of 17).

For each taxpayer, the count of these legitimate dependents was then multiplied by $400 ($500 for 1999 analysis) to calculate the maximum legitimate credit based on available IRS data. [NOTE: The analysis is based on the date of birth information maintained by the IRS and appearing on the RTF.]

This maximum legitimate credit was then compared to the actual amount of CTC allowed (i.e., the ‘Child Tax Credit Amount Computer’ field). Subtracting the actual amount of CTC allowed from the calculated maximum legitimate credit resulted in the non-legitimate/unqualified credit amount of CTC allowed by the IRS. The following table shows the specific numbers from our analysis:

As an example of the methodology, Mr. & Mrs. Taxpayer list 3 dependents on their return, only 2 of which are under the age of 17 (i.e., 2 legitimate dependents). Multiplying the count of 2 legitimate dependents by $400 gives $800 of maximum legitimate credit. If the amount of CTC allowed by the IRS (i.e., amount in the ‘Child Tax Credit Amount Computer’ field on the RTF) is $1,200, then subtracting this amount from the maximum legitimate credit of $800 gives $400 of non-legitimate/unqualified credit.

The results of our analysis are being presented as potential unqualified credits because the IRS maintains only four date of birth fields on its RTF. Therefore, there was no way to electronically verify the ages of more than four dependents. More than 4 dependents were claimed on 240,000 of the 750,000 tax returns. Without reviewing the actual physical tax return there is no way to identify the dependent SSN needed to validate the date of birth for more than 4 dependents. The amount of non-legitimate/unqualified credit associated with these 240,000 tax returns is $128 million. Because of this, we revised our outcome measure and will only claim $211 million for the CTC claims with four or less dependents and revised the number of returns to 512,194.

In addition, the actual number of taxpayers who received potentially unqualified CTCs for 1998 and 1999 was 710,913. This number is less than the number of tax returns (750,000) because some taxpayers received the potentially unqualified CTCs on both the 1998 and 1999 tax returns. We also revised the number of taxpayers we are claiming as outcomes to 505,706 because, as stated above, we could not electronically verify the ages of more than 4 dependents.

Type of Outcome Measure:

Increased Revenue/Revenue Protection - Potential

Value of the Benefit:

By not programming its computers to validate CTC claims, the IRS did not protect over $12 million of revenue due to potentially unqualified ACTCs allowed to approximately 26,000 taxpayers during the period January 1, 1999, to April 6, 2000. However, because we could not electronically verify the date of birth for these returns, we are not claiming this outcome.

Methodology Used to Measure the Reported Benefit:

The $12 million is the actual amount of ACTC allowed (RTF field ‘Additional Child Tax Credit Amount Computer’) to the taxpayers originally identified as having non-legitimate/unqualified regular CTC. The number of taxpayers who received potentially unqualified ACTC was 26,074. This number is less than the number of tax returns (33,000) because some taxpayers received potentially unqualified ACTCs on both the 1998 and 1999 tax returns.

Type of Outcome Measure:

Revenue Protection - Potential

Value of the Benefit:

By modifying its computer programming designed to verify S-SSNs on tax returns, the IRS did not adequately protect revenue from taxpayers who may not have qualified for certain credits and benefits. According to the IRS’ own estimates, approximately 2.7 to 3 million taxpayers have invalid S-SSNs.

Methodology Used to Measure the Reported Benefit:

The 2.7 to 3 million taxpayers with invalid S-SSNs are based on IRS estimates. We did not verify these numbers.

Other Issues Identified During Review

During the course of our review, we provided the Internal Revenue Service (IRS) with detailed information via emails regarding our interim findings and recommendations regarding the need to have better tools to track filing season activities. Management also provided their responses via email. The following are details for these areas and the IRS’ responses.

The process used by IRS personnel to keep track of filing season problems and other activities does not ensure that this tracking will continue when personnel changes are made. The IRS personnel responsible for tracking the Submission Processing Centers’ success in processing tax returns and payments informed us that there are no documented procedures detailing the types of basic tracking techniques required to perform this critical function. For example, during the 2000 Filing Season, an experienced analyst in the office responsible for tracking these activities was assigned to a detail in another office. However, documentation of the analyst’s methods to keep track of volumes for both electronic and paper tax returns was not available for training the temporary replacement.

Informal email messages were the only means used to document the key analyst’ tools for tracking 2000 Filing Season results. The need for written procedures is especially important because of the number of experienced analysts currently detailed to assist in other IRS efforts, including modernization.

Personnel at the IRS agreed that written guidelines were needed to track filing season activities and initiated appropriate corrective action.

The process used to report data for analysis of filing season activities does not provide a high level of assurance that the data are accurate. IRS management relies on a report called the Management Information System for Top Level Executives (MISTLE) as a principal tool for tracking filing season activities. Although we did not perform an in-depth analysis of the overall accuracy of the MISTLE, we did identify instances where incomplete or inaccurate information was contained on the MISTLE. For example, we determined that the Submission Processing Centers did not consistently report the volumes of electronic filed returns they expected to receive and did not always code volume information reported to the National Headquarters for inclusion in reports used to track filing season activities. We determined that for the week ending January 28, 2000, the total number of returns processed was overstated by over 600,000, giving an inaccurate impression that nationwide production was ahead of schedule.

Although spot checks of the report data are conducted, there are no written procedures for routine data validity tests at the National Headquarters. Additionally, IRS guidelines assign responsibility for validating locally generated report data with the Submission Processing Centers. However, no specific guidance is provided to the Centers for validating information prior to submitting it to National Headquarters personnel for inclusion in the MISTLE.

Personnel at the IRS agreed to establish review standards for validating report data.

The IRS efforts to resolve ongoing problems with the equipment used to sort mail did not ensure that envelopes containing checks were easily detected. In a prior Treasury Inspector General for Tax Administration review, we reported that the Service Center Automated Mail Processing System (SCAMPS) check detection accuracy rate needed improving. The SCAMPS is the new mail sorting system for processing the millions of pieces of incoming and outgoing mail handled by IRS Submission Processing Centers. The error rate at that time for several Submission Processing Centers ranged from 55 percent to 88 percent. For a 10-day period ending February 25, 2000, we randomly sampled over 4,000 return envelopes sorted by the SCAMPS as "with remittance," meaning the envelopes contained a check. Our results showed that 228 (approximately 6 percent) of the envelopes contained a remittance. Although this was not a statistically valid sample, it showed that the SCAMPS was still not accurately detecting remittances.

The check detection device on the SCAMPS is very sensitive. In addition to identifying envelopes containing checks, the SCAMPS also identifies and sorts as "with remittance" envelopes containing paper clips, staples, metallic ink on W-2s, and laser copies. Once the SCAMPS sorts the mail into either "with remittance" or "no remittance," IRS employees must physically extract the checks from the envelopes identified by the SCAMPS as "with remittance."

Personnel at the IRS have made efforts to enhance the check detection rate for the SCAMPS, but the rate has not improved. The check detection mechanism is a problem that cannot be corrected.

The IRS has set a goal for depositing checks within 2 days of receipt. Spending extra time opening envelopes without checks could have a significant impact on the IRS’ ability to make timely deposits, in addition to requiring additional staff to open and sort mail during the filing season peak.

The process used to define the wording on mail labels and instructions to taxpayers on where to file their returns did not ensure that envelopes containing checks were easily detected. The IRS developed a dual-label initiative to help improve the SCAMPS check detection accuracy rates. One label was used for returns claiming a refund, the other for returns with no refund. We randomly sampled 1,850 return envelopes that used the non-refund labels (intended to identify envelopes with remittances) for an 11-day period ending February 28, 2000. The SCAMPS sorted these envelopes into the "with remittance" bins based on the label. Our results showed that taxpayers used this label for the following reasons:

1. Enclosed a remittance with return (the intended use) - (45 percent or 838 taxpayers)
2. Applied the refund to the next year’s tax (8 percent or 149 taxpayers)
3. No tax due (10 percent or 186 taxpayers)
4. Other (e.g., balance due but no check in envelope) (less than 1 percent or 9 taxpayers)
5. Refunds (36 percent or 668 taxpayers)

The above results showed that approximately 55 percent of the envelopes we sampled from the "with remittance" bins did not contain a remittance in them. For items 2 through 4 above, taxpayers used the label according to the instructions in their tax packages. However, using the label for these items contradicts the purpose for which the label was intended (i.e., to assist the SCAMPS in detecting remittances).

During our initial discussions with IRS personnel concerning the SCAMPS, we were informed that the decision to use the refund and non-refund wording on the labels was made because the IRS believed using this wording would motivate taxpayers to use the labels (i.e., taxpayers would believe that using the labels helped them get their refunds quicker).

Management at the IRS agreed that the SCAMPS was not accurately detecting remittances and stated they will continue to work with the vendor to improve the accuracy rate. However, they did not agree that the wording on the labels did not meet the intents for which they were designed. The IRS is conducting its own study of the label use and will evaluate the results before making any changes to the wording on the use of the labels. Since these results were early in the 2000 Filing Season, we agree with management’s action to evaluate the results of their study before making any wording changes on the use of the labels.

Overview of Pipeline Processing

When a tax return is received at a Submission Processing Center, it progresses through what the Internal Revenue Service (IRS) calls "Pipeline Processing." This system of operation starts at the loading dock when tax returns are received. From the dock, tax returns are taken to the Receipt and Control area to be run through the Service Center Automated Mail Processing System, which reads the bar-coded envelopes and sorts the tax returns.

After returns are sorted, clerks in the Extraction area open and sort the mail. Tax returns with payments attached go to the Remittance Processing area to have the payments credited to the taxpayers’ accounts and deposited to the United States Treasury. The tax returns are then sent to Code and Edit to be checked for accuracy and prepared for further processing.

All other types of returns, such as balance due and refund, are batched by category and input on the Batch Block Tracking System. The returns are sent to the Code and Edit section to be individually checked for accuracy and completeness and then prepared for further processing.

When returns arrive in the Integrated Submission and Remittance Processing area, data are entered into the computer system; verified; and relayed onto magnetic tape for further processing, math verification, and correction, if necessary.

Finally, magnetic tapes containing tax data are sent to the Martinsburg Computing Center (MCC) for posting to the IRS Masterfile. The MCC generates refund tapes that are sent to the Department of the Treasury Financial Management Services (FMS). The FMS issues refund checks weekly.

The following flowchart provides a graphical overview of Pipeline Processing:

Key Processes

We reviewed judgmental samples of returns from various points in the processing pipeline to determine whether the returns and remittances were processed effectively. Some of the processes we reviewed and our results follow.

Receipt and Control Branch - This branch is the entry point for returns and correspondence received from taxpayers. Employees in Receipt and Control sort the mail, prepare tax payments for deposit, and batch returns and documents. We determined that the returns selected for review were effectively controlled and properly posted to the taxpayers’ accounts.

Remittance Processing Section - The remittance processing section at the Submission Processing Centers is responsible for depositing payments received from taxpayers. Generally, all payments received must be deposited within 24 to 48 hours of receipt. These payments are also credited to the taxpayers’ accounts. Our tests showed that the sample of remittances selected were deposited timely and posted properly to the taxpayers’ accounts.

Code and Edit Section - This section ensures that the correct information from tax documents is identified for subsequent input to the IRS computer system. Work is received from Receipt and Control and processed based on priority. Refund returns are always processed first. Employees review each document for conditions which make it unprocessable, such as missing schedules and supporting forms. They determine whether the return is signed. Employees also review the amounts claimed as deductions or credits that are not allowable by law or reflect some other type of non-compliance. We determined that returns selected were processable and were posted properly to the taxpayers’ accounts.

Error Resolution System (ERS) - This system is used to correct errors made by taxpayers during the initial processing of their tax returns. After data entry operators input information from a tax return into the IRS’ computer system, the computer conducts various checks to verify the accuracy of the information. If the data input does not pass one or more of the checks (math verification, filing status consistent with standard deduction taken, etc.), an error condition is identified. Returns that do not pass these checks are sent to the ERS. Employees review these returns, correct the errors, send the taxpayers appropriate notices if necessary, and send the returns through the pipeline for processing. We determined that IRS employees properly corrected the errors on the sample of returns we selected for review and issued proper notices to taxpayers.

Cycle Proof Listings - These daily listings are used to control returns after they have been correctly input to the IRS’ computer system and filed at storage facilities. Employees compare the returns on the listings to the actual blocks of work to ensure that all returns have been processed through the computer system and are ready to be sent to the storage facilities. We determined that the sample of returns we selected from the storage facilities had been processed properly.

Management’s Initial Response to the Draft Report

The response was removed due to its size. To see the Initial Response, please go to the Adobe PDF version of the report on the TIGTA Public Web Page.

Management’s Amended Response to the Draft Report

Management's Amended Response was removed due to its size. To see the response, please go to the Adobe PDF version of the report on the TIGTA Public Web Page.