The Earned Income Credit Recertification Program Continues to Experience Problems

 

March 2005

 

Reference Number:  2005-40-039

 

 

This report has cleared the Treasury Inspector General for Tax Administration disclosure review process and information determined to be restricted from public release has been redacted from this document.

 

March 11, 2005

 

 

MEMORANDUM FOR COMMISSIONER, SMALL BUSINESS/SELF-EMPLOYED DIVISION
COMMISSIONER, WAGE AND INVESTMENT DIVISION

 

FROM:     Pamela J. Gardiner /s/ Pamela J. Gardiner

                 Deputy Inspector General for Audit

 

SUBJECT:     Final Audit Report - The Earned Income Credit Recertification Program Continues to Experience Problems (Audit # 200340007)

 

This report presents the results of our review of the Earned Income Credit (EIC) Recertification Program.  The overall objective of this review was to evaluate Internal Revenue Service (IRS) corrective actions to recommendations in our December 2000 report on the IRS’ EIC Recertification Program.  Our prior report recommendations were intended to better safeguard Federal Government funds and help ensure the protection of taxpayer rights with the least amount of burden to taxpayers.

In 1997, the Congress implemented legislation to help prevent taxpayers from incorrectly receiving the EIC.  In response, the IRS implemented the EIC Recertification Program in 1999, making taxpayers subject to recertification if they had the EIC denied during an examination.  This denial places recertification indicators on taxpayers’ accounts to help prevent future EIC claims from being accepted until they prove they are eligible to receive the EIC again.  Once taxpayers meet evidentiary requirements, they are “recertified,” and the IRS should remove the recertification indicators.  The number of taxpayers subject to this recertification tripled from 336,000 taxpayers as of September 1999 to almost 1 million taxpayers as of December 2003.

This report includes an evaluation of the EIC Recertification Program as it relates to the effectiveness of the IRS corrective actions to our prior report recommendations.  The conditions identified in our prior report included not accurately removing recertification indicators, not timely releasing suspended refunds, not clearly and accurately communicating with taxpayers, and the need to change certain processing procedures and the recertification criteria for certain taxpayers.  We did not review those processes and procedures that we previously reported were working as intended, so we do not comment on them in this report.  Also, while some of the analysis and evaluations of corrective actions in this report are several months old, we believe the conditions and issues we identified are still valid and the recommendations we make are still relevant to the current environment.

The IRS has taken some corrective actions to improve the EIC Recertification Program.  For example, it has developed monthly reports to identify accounts needing the recertification indicators removed or refunds released, performed “clean-ups” to remove indicators that should have been removed previously or to release refunds, and implemented programming to automatically remove indicators or release refunds based on certain Examination function closing actions.  The IRS also implemented programming to help prevent unnecessary examinations and adopted our prior recommendation to change the recertification of certain taxpayers.   

While the corrective actions did have some positive impact, many were ineffective, incomplete, or inaccurate.  Additionally, the IRS did not specifically address some of our prior recommendations.  As a result, we believe the conditions and problems previously identified have not improved, leading to taxpayers not receiving the EIC; delayed refunds; increased taxpayer burden for taxpayers forced to recertify more than 1 time; wasted IRS resources; and a loss of Federal Government funds, including approximately $110 million of EIC allowed without taxpayers demonstrating they were entitled to receive the EIC.  We believe these conditions also cause taxpayer confusion about the EIC Recertification Program, which could result in taxpayers not claiming the EIC they are entitled to receive or not questioning incorrect IRS denials of the EIC.  Not receiving the EIC could then subject these low-income taxpayers to financial hardship. 

The IRS does not have a consistent process for ensuring recertification indicators are accurately removed.  We estimate that, as of April 2003, 52,000 taxpayers did not have recertification indicators removed from their accounts when the EIC was allowed; over 10,000 of them had almost $21 million of EIC improperly denied.  We also estimate approximately 14,800 taxpayers had the recertification indicators incorrectly removed from their accounts or were allowed approximately $26.5 million of EIC without demonstrating they were entitled to it.

The IRS had not timely released approximately $4 million in incorrectly suspended refunds for over 3,200 taxpayers.  These refunds had been incorrectly suspended for an average of over 1 year.  After we referred this problem to the IRS, additional delays of 1 year occurred to release approximately 1,000 of these refunds.

The IRS sent taxpayers approximately 850,000 letters or electronic messages that did not clearly or accurately communicate the EIC recertification requirements, did not specifically inform taxpayers they were recertified, or referred to an enclosed recertification tax form that was not in fact enclosed.  This made it harder for taxpayers to understand and comply with recertification requirements.    

The IRS had significant problems trying to implement changes regarding when taxpayers are recertified.  For example, the IRS provided inconsistent information to taxpayers and IRS employees as to what tax years would be affected by the changes.  Additionally, programming problems caused some taxpayers to have their EIC incorrectly denied, while allowing approximately 36,000 others $84 million of EIC without their having to demonstrate they were entitled to receive it.

We recommended the Commissioner, Wage and Investment Division, develop clear, consistent, and comprehensive guidelines explaining when taxpayers are recertified and ensure employees are consistently applying the guidelines.  Requests for computer programming should be more detailed, and programming should be sufficiently tested before being implemented.  Required quality reviews should be performed and the results evaluated for any necessary corrective actions.  Inventory reports should be modified and evaluated to help ensure Examination closing actions are appropriate.  Problem Correction Reports should identify unresolved suspended refunds and be worked timely.  Communications to taxpayers should be further revised to inform them when they are recertified and to notify potentially eligible taxpayers about the income-only EIC.  The communications should also clearly explain why the EIC is not being allowed, that filing an Information to Claim Earned Income Credit After Disallowance (Form 8862) does not by itself recertify taxpayers, and that filing Form 8862 will likely result in an examination.  Additionally, the IRS should include Form 8862 with the letters that deny the EIC because Form 8862 was not filed.  Finally, existing computer programming should be corrected to prevent certain electronically filed returns claiming the EIC from being incorrectly rejected.

Management’s Response:  IRS management agreed with our recommendations and has initiated corrective actions.  Examples include:

  • Updating the Internal Revenue Manual to clearly explain the criteria that causes the automated release of suspended taxpayer refunds and developing a job aide that will remind tax examiners to remove recertification indicators and release frozen refunds when automation will not.
  • Programming changes to correct the Problem Correction Reports and make them more manageable. 
  • Sending clear letters to taxpayers informing them of the potential for EIC without a qualifying child when a recertification indicator is on their account and informing them when they have successfully recertified for the EIC.
  • Revising communications to taxpayers to include language on the recertification process and the Form 8862.

IRS management also agreed with our outcome measures, with the exception of the $20.8 million (reported as “almost $21 million”) in EIC improperly denied.  Management believes there was too wide a range of variability caused by the precision level used to calculate this measure, and therefore they do not agree with the figure we report.  However, they do acknowledge that the IRS improperly denied the EIC by auditing taxpayers whose recertification indicator should have been removed.  Management’s complete response to the draft report is included as Appendix V.

Office of Audit Comment:  While we acknowledge a wide range of variability in the calculation of this outcome measure, we believe the $20.8 million figure to be an appropriate measure of the potential impact of improperly denying the EIC to taxpayers.  Instead of using the precision level stated in the report, which provided an outcome measure as high as $84 million, we used the average amount of the child-related EIC allowed by the IRS for tax returns processed during Calendar Years 2001 to 2003 times the number of projected cases where the EIC was improperly denied to arrive at the $20.8 million figure.  In addition, the $20.8 million is conservative because our sample was taken at a “point in time” and there would have been additional taxpayers who had the EIC improperly denied that would not have been included in our calculation. 

Copies of this report are also being sent to the IRS managers affected by the report recommendations.  Please contact me at (202) 622-6510 if you have questions or Michael R. Phillips, Assistant Inspector General for Audit (Wage and Investment Income Programs), at (202) 927-0597.

 

Table of Contents

Background

Corrective Actions Were Taken to Improve the Recertification Program

Recertification Indicators on Taxpayers’ Accounts Are Not Always Accurate

Recommendation 1:

Recommendation 2:

Recommendation 3:

Taxpayers Did Not Always Timely Receive Refunds

Recommendation 4:

Communications With Taxpayers Do Not Fully Explain the Recertification Requirements

Recommendation 5:

Changes Regarding When Taxpayers Are Recertified Were Not Effectively Implemented

Recommendation 6:

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 – Management’s Response to the Draft Report

 

Background

The Congress established the Earned Income Credit (EIC) to help alleviate poverty and provide work incentives to low-income taxpayers.  For Calendar Year (CY) 2003, over 21 million taxpayers received the EIC, totaling almost $37 billion.

The most recent Internal Revenue Service (IRS) study of EIC compliance estimated that between $8.5 billion and $9.9 billion of the EIC claims filed for Tax Year (TY) 1999 should not have been paid to taxpayers.  To help address the EIC compliance problems, the Congress passed legislation requiring taxpayers that had the EIC denied during examinations to prove eligibility before receiving the EIC again.  In response to this legislation, the IRS implemented the EIC Recertification Program in 1999.  The number of taxpayers subject to recertification tripled from 336,000 taxpayers as of September 1999 to almost 1 million taxpayers as of December 2003.

There are two ways taxpayers qualify for the EIC, and the amount allowed is dependent on the income level, the filing status, and the number of qualifying children claimed.

·         Child-Related EIC – taxpayers claiming one or two qualifying children.  For TY 2003, it ranges from $1 to approximately $4,200 and reaches zero when earned income reaches approximately $34,700 for married filing jointly taxpayers and $33,700 for other taxpayers.

·         Income-Only EIC – taxpayers that meet age requirements but do not claim children for EIC purposes.  For TY 2003, it ranges from $1 to approximately $380 and reaches zero when earned income reaches $12,230 for married filing jointly taxpayers and $11,230 for other taxpayers.

The recertification process

The recertification requirement begins after the IRS examines a taxpayer’s tax return and denies the EIC claimed, which places a recertification indicator on the taxpayer’s account.  This indicator helps prevent, although it does not ensure, the taxpayer from receiving the EIC again until he or she demonstrates eligibility to receive it.  Once the taxpayer meets the evidentiary requirements to prove eligibility, he or she is “recertified,” and the IRS should remove the recertification indicator. 

There are two common methods to begin the recertification process.

·         The taxpayer can request an examination reconsideration to reevaluate the examination results that denied the EIC.  If the IRS then allows the EIC previously denied, the taxpayer is recertified.

·         The taxpayer can file an Information to Claim Earned Income Credit After Disallowance (Form 8862) with the next tax return that claims the child-related EIC.  This alerts the IRS that the taxpayer wants to begin the recertification process.  The IRS then determines whether to examine the taxpayer.

If an examination is conducted, the IRS requests evidentiary documentation from the taxpayer to determine if he or she should be recertified.  Generally, a tax return with a Form 8862 should be examined unless (1) the taxpayer had previously met recertification requirements, but the IRS had not removed the recertification indicator, or (2) the taxpayer is serving in a combat zone.

In December 2000, we issued a report on the IRS’ EIC Recertification Program that included recommendations to help improve the Program.  Our current report details our evaluation of the effectiveness of the IRS corrective actions to our prior recommendations.  We did not review those processes and procedures that we previously reported were working as intended, so we do not comment on them in this report.  Also, while some of the analysis and evaluations of corrective actions in this report are several months old, we believe the conditions and issues we identified are still valid and the recommendations we make are still relevant to the current environment.

Our current audit was performed at the Compliance function of the Wage and Investment (W&I) Division Headquarters in Atlanta, Georgia, and at the Fresno, California, and Austin, Texas, Compliance Sites during the period May 2003 through May 2004.  The audit was conducted in accordance with Government Auditing Standards.  Detailed information on our audit objective, scope, and methodology is presented in Appendix I.  Major contributors to the report are listed in Appendix II.

Corrective Actions Were Taken to Improve the Recertification Program

The IRS took actions to address some of the conditions discussed in our prior report.  Some specific actions to address our recommendations included the following.

·         Recommendation:  Ensure the IRS accurately removes recertification indicators.  Corrective Actions:  The IRS established monthly Problem Correction Reports to identify accounts needing the indicators removed, performed a clean-up to remove indicators that should have been removed previously, implemented programming to automatically remove indicators based on certain Examination function closures, and revised processing procedures.  The IRS also implemented “filter programming” to help prevent tax returns from being examined if the IRS had previously allowed the EIC but did not remove the indicators.

·         Recommendation:  Timely release suspended refunds.  Corrective Actions:  The IRS performed a clean-up to identify refunds that needed to be released and developed monthly Problem Correction Reports to identify any subsequent accounts needing refunds released.  The IRS also revised its computer programs to immediately release the portion of the refunds not related to the questioned EIC claims, and to automatically release any remaining refunds when Examination function actions were complete. 

·         Recommendation:  Clearly and accurately communicate with taxpayers.  Corrective Action:  The IRS revised the letter that informs taxpayers their EIC is being denied during an examination and they are now subject to recertification.  

·         Recommendation:  Instruct employees as to what is needed to prove taxpayers are qualified to receive the income-only EIC.  Corrective Action:  The IRS revised processing procedures to specifically instruct employees as to what information and actions are needed to prove taxpayers are qualified to receive the income-only EIC during examinations that deny the child-related EIC.

·         Recommendation:  Change the EIC Recertification Program to help ensure taxpayers are recertified for the reason the EIC was denied.  Corrective Action:  The IRS changed the EIC Recertification Program so taxpayers denied the child-related EIC will not be considered to be recertified simply by claiming and receiving the income-only EIC.

While the IRS’ corrective actions did have some positive impact, many of the actions were ineffective, incomplete, or inaccurate.  Additionally, the IRS did not specifically address some of our prior recommendations. 

Ineffective corrective actions led to taxpayers not receiving the EIC or not timely receiving refunds, taxpayer burden in the form of unnecessary examinations or time and costs preparing unnecessary forms, unclear or confusing communications to taxpayers that made it difficult for them to understand and comply with the EIC recertification requirements, wasted IRS resources, and a loss of Federal Government funds.  In addition, taxpayers may become confused and give up on the recertification process and not receive their EIC.  The confusion issue was reported to the IRS in an April 2002 Government Accountability Office report.

While we recognize that the population of taxpayers in the recertification program has tripled since our prior report, we believe that the conditions and problems previously identified have not improved.  For example:

·         The estimated number of taxpayer accounts that did not have recertification indicators accurately removed increased from 15,500 to almost 64,000.

·         The average refund delay increased from 9 weeks identified in our prior report to over 1 year for over 3,200 taxpayers.

·         The number of inadequate IRS communications to taxpayers increased from approximately 400,000 to 850,000. 

Recertification Indicators on Taxpayers’ Accounts Are Not Always Accurate

Having accurate recertification indicators is crucial to the success of the EIC Recertification Program.  However, the IRS does not have a consistent process for ensuring recertification indicators are accurately removed.  The IRS did not properly or timely remove recertification indicators for some taxpayers who had been recertified, while other taxpayers had the indicators improperly removed from their accounts.

Not removing indicators caused some taxpayers to have their EIC incorrectly denied during original tax return processing or during examinations.  For example, the EIC can be denied when taxpayers become confused and simply accept IRS decisions to deny the EIC or when IRS employees make incorrect decisions to deny the EIC.  Additionally, unnecessary examinations forced some taxpayers to recertify more than one time, causing taxpayer burden and wasting IRS resources.  Improper removal of indicators allowed some taxpayers to receive the EIC without having to demonstrate they were entitled to receive it.

Recertification indicators were not properly or timely removed from taxpayer accounts

We estimate that, as of April 2003, 52,000 taxpayers still had recertification indicators on their accounts even though the IRS had allowed subsequent EIC claims and the indicators should have been removed.  While subsequent IRS actions corrected many of these accounts, we estimate 17,000 of these taxpayers still had the recertification indicators on their accounts as of March 2004, almost a year later.  Additionally, of the 52,000 taxpayers, we estimate that 10,500 had EIC claims of almost $21 million improperly denied and 17,000 had their EIC claims unnecessarily examined.

Recertification indicators were improperly removed from taxpayer accounts

We estimate the IRS improperly removed recertification indicators from 11,900 taxpayer accounts between January 2002 and August 2003.  The improper removal of the recertification indicators allowed taxpayers to receive almost $20 million of EIC without proving they were entitled to receive it.  

We define “improper removal” as removing indicators without the taxpayers demonstrating they were eligible to receive the EIC again.  This included either of the following situations:

·         The IRS had not allowed the EIC during a tax year subject to recertification.

·         IRS Examination function employees made incorrect decisions when closing returns without examinations (commonly referred to as nonexamined closures).  While the decision not to examine the return is incorrect, allowing the EIC without examining the return still recertifies the taxpayers and the recertification indicators should be removed.

Programming to help prevent unnecessary examinations was not fully effective

The IRS implemented “filter programming” in January 2003 to help prevent returns from being unnecessarily examined for those taxpayers previously allowed the EIC without having recertification indicators removed.  This programming helped prevent some taxpayers from having to recertify more than once and helped prevent the EIC from being improperly denied during examinations.

Although the filter programming identified approximately 14,000 returns for the period January 2003 through early August 2003, we identified problems affecting approximately 5,000 of the 14,000 returns.

·         Approximately 2,200 tax returns should have been sent to the Examination function for examination but were not.  Specifically, the filter programming incorrectly prevented the examinations.  These 2,200 taxpayers received an estimated $5.1 million of EIC without having to demonstrate eligibility to receive the EIC again.

·         Approximately 2,000 tax returns processed during January and February 2003 were unnecessarily examined, with over 900 taxpayers having their EIC claims incorrectly denied.  The IRS identified the 2,000 returns using listings based on revisions to filter programming in February 2003.  The IRS’ intent was to not examine these returns; however, the returns were examined, in part because the IRS did not provide adequate instructions for processing the listings.

·         Approximately 700 tax returns, processed in January and February 2003 and claiming an estimated $1.6 million of EIC, were going to be closed without examinations due to programming problems.  After we advised the IRS of the problems, it revised the filter programming and advised the Compliance sites that these 700 returns needed to be examined.

Causes for inaccurate indicators

Some causes for the incorrect recertification indicators are similar to those identified in our prior report.  Additionally, although the IRS took corrective actions based on our prior recommendations, the actions taken were not effective.

·         The IRS did not develop clear, consistent, and comprehensive guidelines for determining when taxpayers are actually recertified, which caused confusion among IRS employees.

·         The programming to generate monthly Problem Correction Reports, implemented in March 2002, identified only a small percentage of the accounts needing the recertification indicators removed.  The programming did not include relevant information that we provided in our prior report to help identify accounts needing indicators removed.  Additionally, the IRS did not revise the programming in CY 2003 to reflect changes for processing recertification-related tax returns.

·         The July 2003 clean-up programming removed recertification indicators for approximately 25,000 taxpayers.  However, the clean-up identified less than one-half of the accounts needing the indicators removed and incorrectly removed approximately 5,000 of the 25,000 indicators.  Based on our referral, the IRS partially corrected the programming and removed over 9,000 additional indicators.

·         The requests to develop recertification programming did not have sufficient details to ensure the programming was complete and accurate.  In addition, programming was not sufficiently tested before being implemented.

·         Employees outside the Examination function allowed the EIC for some taxpayers subject to recertification without obtaining approval from the Examination function.  The IRS’ intent is to have Examination function employees make the decisions as to whether to allow the EIC for taxpayers subject to recertification.  This helps ensure the decisions to allow the EIC are correct and the recertification indicators are accurately removed.  However, the IRS did not develop controls to ensure Examination function employee involvement in the decisions to allow the EIC.

·         The IRS did not adequately analyze inventory reports to identify incorrect decisions by Examination function employees to allow the EIC without examining the related tax returns (nonexamined closures).  Additionally, the inventory reports did not contain information that would have made it easier to identify nonexamined closure problems.  For example, the inventory reports included only one type of nonexamined closure, and this one type did not have separate volumes for each EIC Program.

·         The IRS did not always perform, or adequately evaluate, the required quality review of nonexamined closures.

Recent Management Actions:  The IRS designed new programming for CY 2004 to automatically remove the recertification indicators based on certain Examination function closures.  The IRS expects this programming change will impact a large percentage of the Examination function cases and should significantly reduce the number of indicators that are not properly removed.  However, our evaluation of the programming showed problems that would incorrectly remove indicators for some taxpayers while not removing indicators for other taxpayers.

Based on our referral, IRS management advised us they had corrected some of the programming problems we identified, but several others remained.  Due to time constraints, we were unable to evaluate the overall effectiveness of the final programming.

Recommendations

To help ensure EIC recertification indicators on taxpayer accounts are accurate and taxpayers receive the EIC only when they are entitled to it, the Commissioner, W&I Division, should:

1.      Develop clear, consistent, and comprehensive guidelines explaining when taxpayers are recertified and ensure the guidelines are consistently followed.  Recertification guidelines should explain the EIC Recertification Program complexities and nuances so managers, analysts, programmers, and other individuals will be able to know exactly how the Program should be implemented.  This will allow IRS employees to know what should be included in processing procedures, letters to taxpayers, tax publications, tax packages, and tax forms. 

Management’s Response:  The IRS plans to update the Internal Revenue Manual (IRM) to include clear language explaining recent computer programming changes and the criteria that causes the automated release of suspended taxpayer refunds.  The IRS will also develop a job aide that will remind tax examiners to remove recertification indicators and release frozen refunds when automation will not.

IRS management disagreed with the $20.8 million (reported above as “almost $21 million”) in EIC improperly denied.  Management believes there was too wide a range of variability caused by the precision level used to calculate this measure, and therefore they do not agree with the figure we report.  However, they do acknowledge that the IRS improperly denied the EIC by auditing taxpayers whose recertification indicator should have been removed. 

Office of Audit Comment:  While we acknowledge a wide range of variability in the calculation of this outcome measure, we believe the $20.8 million figure to be an appropriate measure of the potential impact of improperly denying the EIC to taxpayers.  Instead of using the precision level stated in the report, which provided an outcome measure as high as $84 million, we used the average amount of the child-related EIC allowed by the IRS for tax returns processed during CYs 2001 to 2003 times the number of projected cases where the EIC was improperly denied to arrive at the $20.8 million figure.  In addition, the $20.8 million is conservative because our sample was taken at a “point in time” and there would have been additional taxpayers who had the EIC improperly denied that would not have been included in our calculation. 

2.      Correct existing EIC Recertification Program computer programming and ensure future requests for programming are of sufficient detail to ensure applicable IRS employees know exactly what is needed.  The computer programming should be sufficiently tested to identify and correct potential problems prior to implementation.  In addition, computer programming should ensure the EIC allowed for recertification cases reflects the approval of Examination function employees.

Management’s Response:  The IRS completed testing in early December 2004 on a current computer programming request to correct Problem Correction Reports and make them more manageable.  For future programming requests, analysts and programmers will work closely to ensure testing and delivery of desired functionality. 

3.      Ensure required quality reviews of nonexamined closures are performed, the results are evaluated, and corrective actions are taken if appropriate.  The IRS should change Examination function inventory reports to include counts for each type of nonexamined closure for each EIC Program.  These inventory reports should be evaluated to help ensure nonexamined closures are appropriate.

Management’s Response:  The IRM requires managerial approval of all nonexamined closures, with the exception of those cases closed “accepted as filed” by the Classification function.  The IRS requested this exception to eliminate the burden placed on managers to review and approve such closures when significant volumes generate at one time.  However, quality reviews were made of a percentage of these cases.  Since that time, systemic programming changes have been implemented to reduce the volume of these closures from 28,580 in Fiscal Year 2002 to 1,365 in Fiscal Year 2004 across all 5 W&I Division sites.

The IRS does not plan to change the inventory reports to include counts for each type of nonexamined closure since the current Audit Inventory Management System reports stratify the closures by project code.  These reports are reviewed and monitored.  However, Headquarters Examination and Quality Analysts will work together to review the quality review procedures and the results of the reviews performed to determine if any changes should be made.

Taxpayers Did Not Always Timely Receive Refunds

The IRS did not always release refunds timely.  The IRS suspends refunds when taxpayers subject to recertification file tax returns with Forms 8862.  When Examination function processing actions are complete, any refunds due the taxpayers should be released.  Beginning with CY 2004 processing, the IRS suspends only that portion of the refund related to the total EIC claimed.  Prior to 2004, the IRS suspended the entire refund.

We reported this same issue in our December 2000 report and stated that refund delays occurred in part because IRS procedures did not provide time periods for releasing suspended refunds.  To address this issue, the IRS implemented programming in March 2002 to generate monthly Problem Correction Reports to identify accounts whose refunds had not been released.  The IRS response to our prior report stated that the IRS planned to provide instructions to employees to work the Problem Correction Reports within 5 business days.  The IRS also planned a clean-up to identify accounts needing to have the refunds released.

However, during CY 2003, we identified over 3,200 taxpayer accounts with refunds totaling over $3.9 million that were not timely released, with delays averaging over 1 year.  After we referred this problem to the IRS, additional delays of 1 year or more occurred to release approximately 1,000 of the 3,200 refunds. 

Overall, delays occurred because:

·         The Problem Correction Report programming did not identify some accounts needing to have refunds released.

·         The Problem Correction Reports at 4 of 10 Compliance sites were not worked at all, and 2 sites did not work the reports timely.  Also, the IRS did not evaluate monthly summary reports that clearly showed some Compliance sites were not working the Problem Correction Reports.

·         The processing procedures at the time of our review did not include instructions to work the Problem Correction Reports within 5 business days.

·         The clean-up did not identify or otherwise resolve all accounts needing to have refunds released.

Recent Management Actions:  In addition to the programming change that allows the IRS to immediately release that portion of the refund not related to the EIC claim, the IRS also implemented programming changes in CY 2004 to automatically release any remaining refunds when Examination function actions were complete.  While we were unable to test the programming during our audit, the IRS has stated that the programming was working as intended.

Recommendation

To further ensure taxpayers timely receive EIC refunds affected by the recertification process, the Commissioner, W&I Division, should:

4.      Ensure Problem Correction Reports identify unresolved suspended refunds, issue procedures for completing actions for the Problem Correction Reports within specified time periods, and evaluate monthly summary reports to assure the Problem Correction Reports are worked timely.  

Management’s Response:  As detailed in the response to Recommendation 2, the programming to fix the reports was tested and the corrected output was distributed beginning on January 4, 2005.  Per the IRM, the Examination function has 5 days to work the report by annotating actions taken.  The Examination function is required to maintain these listings for 6 months.

Communications With Taxpayers Do Not Fully Explain the Recertification Requirements

One of the IRS Commissioner’s priorities is to make it easier for taxpayers to understand and comply with the EIC tax law.  However, a number of IRS communications sent to taxpayers regarding the EIC Recertification Program are incomplete, inaccurate, or do not clearly explain the recertification process.  These communications included approximately 850,000 letters or electronic messages sent to taxpayers as discussed below. 

We believe that unclear, confusing, or inaccurate communications can make it difficult for taxpayers to understand and comply with the EIC recertification requirements and that a significant number of taxpayers would have been confused by these communications.  This may cause taxpayers to give up on the recertification process and not receive their EIC.

Communications did not contain specific information to assist taxpayers

The IRS did not specifically inform taxpayers that they were recertified or that potentially qualified taxpayers subject to recertification might be entitled to the income-only EIC.

·         Over 100,000 taxpayers had recertification indicators removed during CYs 2002 or 2003 without being specifically informed they were recertified and did not need to subsequently file Form 8862.  This contributed in part to over 50,000 taxpayers unnecessarily filing Forms 8862 in CY 2003.

Specific notification would help prevent taxpayers from spending time and/or money submitting unnecessary Forms 8862, help prevent taxpayer confusion, and help ensure taxpayers receive the EIC.  It would also provide documentation to help taxpayers resolve examinations or EIC denial letters incorrectly caused when IRS employees do not remove recertification indicators.

·         If taxpayers do not claim the income-only EIC, the IRS normally sends a letter to notify potentially eligible taxpayers that they might be entitled to receive it.  However, these letters are not sent to taxpayers who are subject to recertification. 

We stated in our prior report that over 6,000 taxpayers subject to recertification did not receive the letter for the first 9 months of CY 1999.  Although we do not have estimates for subsequent years, based on IRS statistics that over 40 percent of taxpayers sent this letter respond and receive the income-only EIC, we believe a significant number of eligible taxpayers subject to recertification did not receive the income-only EIC.

One type of letter was incomplete

One type of letter used to deny the EIC during tax return processing was incomplete.  Specifically, the letter stated Form 8862 was being included with the letter, but the Form was not included.  Over 200,000 taxpayers did not receive the Forms 8862 as enclosures during CYs 2001 through 2003.  Taxpayers needing Forms 8862 then had to otherwise obtain the Forms.

One type of letter was inaccurate and unclear

The IRS sends letters advising taxpayers that they are subject to recertification.  The IRS sent over 370,000 of these letters in CY 2003.

·         The letter had two inaccurate headings as to when Forms 8862 should be filed.  The IRS fixed the inaccurate headings in June 2003 after our referral.  If taxpayers followed the inaccurate instructions, they would have had their EIC denied during tax return processing.

·         The IRS response to our prior report stated that the IRS revised the letter to tell taxpayers they may qualify for the income-only EIC even though the child-related EIC was denied.  However, the letter was later changed and the revised wording that constituted the corrective action was deleted.

Two types of communications did not clearly explain the recertification process

Our current audit identified additional problems not included in our prior report.  Specifically, taxpayers subject to recertification who claim the EIC but do not file Forms 8862 with their original tax returns will receive one of two types of IRS communications.  These include a reject message when the IRS rejects electronically filed returns or a letter when the EIC is denied for paper returns.  For CY 2003, this included over 176,000 reject messages and over 35,000 letters.  However, these two types of communications do not clearly explain the recertification process.

·         The communications do not adequately explain why the EIC is not being allowed.  Specifically, the IRS does not explain that the EIC was previously denied during an examination and Form 8862 is needed to begin the recertification process.

·         The communications may lead some taxpayers to believe simply filing Form 8862 will recertify them.  For example, the reject message states, “IRS Master File indicates that the taxpayer must file Form 8862 to claim [the] Earned Income Credit after disallowance.”  This message encourages taxpayers to resubmit returns with Forms 8862 without regard to whether the taxpayers believe they have documentation to prove EIC eligibility.  Although the letter to taxpayers filing paper returns does state that additional information might be requested, the wording still implies the Form 8862 recertifies the taxpayer.

·         The communications do not explain that filing Form 8862 will likely result in an examination, with taxpayers being asked to provide additional information.  A comparison shows the reject message for electronic returns does not state that additional information may be requested, while the letter for paper returns states that additional information “might” be requested.

For those taxpayers who file Forms 8862 believing this by itself recertifies them, the following adverse situations could result:

·         IRS examination procedures could subject the taxpayers to a 2-year ban for receiving the EIC due to reckless or intentional disregard of the EIC rules.  For example, the IRS may apply the 2-year ban if taxpayers subject to recertification claim the EIC again without being able to provide sufficient documentation during an examination.

·         The IRS could unnecessarily expend resources to examine returns for taxpayers who did not know they needed documentation to prove EIC eligibility.

Causes for the communication problems

We identified several causes for the IRS communication problems.

·         The IRS did not adopt our prior report recommendation to specifically advise taxpayers they are recertified.  In addition, corrective actions were incomplete to notify potentially eligible taxpayers subject to recertification that they might be eligible for the income-only EIC.

·         The IRS had advised us during our prior audit that corrective action had been taken to provide Form 8862 as an enclosure, but the IRS did not ensure the corrective action was implemented.

·         The inaccurate information in letters advising taxpayers they are subject to recertification occurred because the IRS had not clearly defined which tax year would be the first year for which Form 8862 would not be needed to receive the income-only EIC.  This issue is also a cause for problems discussed further in the next section of this report.

Recent Management Actions:  The IRS stated it has revised Publication 596 (Earned Income Credit (EIC)) to provide clearer information about the recertification process.  The Publication provides instructions explaining when taxpayers are required or not required to provide Form 8862, and advises taxpayers that additional documents may be required to verify EIC even if Form 8862 is submitted.  In addition, the IRS revised the Form 8862 to make it clearer and to reduce the time needed to complete the Form, and developed a new EIC for Taxpayers Without a Qualifying Child (Form 886-H-EIC) to help taxpayers determine if they qualify for income-only EIC.

Recommendation

To further ensure taxpayers are properly and accurately notified about their involvement in the EIC Recertification Program and information provided to the taxpayers after filing their return is complete and understandable, the Commissioner, W&I Division, should:

5.      Further revise communications to taxpayers to specifically inform them when they are recertified and notify potentially eligible taxpayers subject to recertification that they may still be entitled to the income-only EIC.  The IRS should also revise communications to clearly explain why the EIC is not being allowed, that filing Form 8862 does not by itself recertify taxpayers, and that filing Form 8862 will likely result in an examination.  Additionally, the IRS should include Form 8862 with the letters that deny the EIC because Form 8862 was not filed.

Management’s Response:  The IRS requested computer programming changes to send letters to taxpayers informing them of the potential for EIC without a qualifying child when a recertification indicator is on their account and informing them when they have successfully recertified for the EIC.  In addition, communications were revised or are being revised to include language on the recertification process and include the Form 8862.

Changes Regarding When Taxpayers Are Recertified Were Not Effectively Implemented

During our prior review, simply filing a Form 8862 and receiving the income-only EIC for less than $350 would recertify taxpayers.  These taxpayers could then receive up to $3,800 of the child-related EIC without proving eligibility to receive it.  This situation did not ensure only eligible taxpayers received the EIC, so we recommended the IRS change the guidelines for the EIC Recertification Program regarding when taxpayers are recertified.

The IRS adopted our recommendation and issued new regulations in June 2001.  Under the new regulations, as they relate to taxpayers denied the child-related EIC during examinations:

·         Taxpayers have to file a Form 8862 with tax returns that claim the child-related EIC and demonstrate they are entitled to the EIC before being recertified.

·         Taxpayers do not have to file a Form 8862 to receive the income-only EIC, but these taxpayers would not be considered recertified.

While the new regulations prevented over 40,000 taxpayers from being recertified simply because the taxpayers received the income-only EIC during CY 2003, the IRS had significant difficulties implementing the regulations.  Specifically, the IRS provided inconsistent information to taxpayers and IRS employees and experienced three programming problems for CY 2003 processing.  These programming problems resulted in loss of Federal Government funds, denial of taxpayer rights and entitlements, and increased taxpayer burden.

Inconsistent information

The IRS provided inconsistent information to taxpayers and IRS employees as to which tax years would be affected by the new regulations.  We compared the letters sent to taxpayers denying the EIC, the Form 8862 Instructions, and the IRS processing procedures.  Each of these documents differed as to the first tax year that Form 8862 would not be required to receive the income-only EIC.  The first tax year included TYs 2001, 2002, and 2003, respectively. 

The inconsistent information occurred because the IRS did not clearly define which tax year would be the first for which Form 8862 would not be needed to receive the income-only EIC.  After we raised this issue, the IRS decided TY 2002 would be the first year that Forms 8862 would not be required for taxpayers claiming the income-only EIC.  We believe the inconsistent information would have confused IRS employees and taxpayers.

Incorrect programming

Not having a clear definition of which tax year would be the first year for which Form 8862 would not be needed for the income-only EIC contributed to three programming problems.  This allowed taxpayers to incorrectly receive a significant amount of the EIC.  Additionally, this caused some taxpayers to not receive the income-only EIC, to have their refunds delayed, or to expend time and money submitting unnecessary Forms 8862. 

·         The IRS incorrectly denied income-only EIC totaling over $400,000 for more than 2,100 taxpayers who filed paper TY 2002 returns without Forms 8862.  We advised the IRS of the related programming problem in February 2003 during the planning phase of this review.  Although the IRS took timely action allowing the EIC for over 900 taxpayers, delays of more than 1 year have occurred for allowing the EIC for the remaining 1,200 taxpayers.  Programming changes resulting from our referral did allow correct processing for over 21,000 taxpayers to receive the income-only EIC amounting to over $4.7 million during CY 2003.

·         The IRS incorrectly allowed $84 million in child-related EIC claimed by approximately 36,000 taxpayers who filed TY 2002 paper returns.  These taxpayers received the EIC without filing a Form 8862 and without having to provide documentation during examinations proving that they were actually entitled to the EIC.  When we advised the IRS of this programming problem, we were informed the IRS had identified the problem 1 week earlier and the problem occurred while fixing the programming condition described above for the income-only EIC.

·         The IRS incorrectly rejected electronically filed TY 2002 returns claiming the income-only EIC without a Form 8862.  Although we do not know the number affected, we believe some taxpayers would not have received the EIC because they would have been confused and would therefore not have resubmitted rejected returns.

Recommendation

The IRS took corrective action for the first two programming problems discussed above; the third programming problem had not been corrected at the time of this audit.  Consequently, to ensure taxpayers are not required to file unnecessary Forms 8862 and taxpayers receive the EIC, the Commissioner, W&I Division, should:

6.      Change computer programming, where appropriate, so electronically filed returns claiming the income-only EIC are not rejected because Forms 8862 are not filed.

Management’s Response:  The IRS is revising programming to not reject electronically filed individual tax returns that claim the EIC without qualifying children (income-only EIC) and without a Form 8862 attached when an EIC recertification indicator is present.

 

Appendix I

 

Detailed Objective, Scope, and Methodology

 

The overall objective of this review was to evaluate Internal Revenue Service (IRS) corrective actions to recommendations in our first audit report on the Earned Income Credit (EIC) Recertification Program.  The recommendations made in that report were intended to better safeguard Federal Government funds and help ensure the protection of taxpayer rights with the least amount of burden to taxpayers.  Due to time constraints and because the prior findings that resulted in two recommendations had been addressed by the Government Accountability Office and by one of our other reports, our current audit evaluated corrective actions on only 7 of 12 prior report recommendations.

Recertification indicators have several “values,” depending on why the EIC is being denied.  Our audit covered only value “1,” which principally includes the child-related EIC denied for other than reckless or intentional disregard of the EIC rules or fraud.  Value 1 covers by far the largest number of taxpayers subject to recertification and had grown from 336,000 taxpayers as of September 1999 to almost 1 million as of December 2003.  To accomplish our objective, we:

I.                   Evaluated the corrective action to Recommendation 1 in our December 2000 report regarding accurately removing recertification indicators.

A.    Determined whether recertification indicators were released when the IRS allowed the EIC claimed in a recertification tax year.  This included reviewing 5 separate random samples totaling 1,100 taxpayer accounts and evaluating various IRS programming.  We used random samples to be able to project to the related populations the number of “exception cases” and the related dollar amounts.  See Appendix IV for sample selection information.

B.     Evaluated the Calendar Year (CY) 2003 “filter programming” designed to prevent examinations of tax returns on which the EIC had been previously allowed without removing the recertification indicators.  Our methodology would identify returns that should not have been filtered (examinations were incorrectly prevented) but would not identify returns that should have been but were not filtered (examinations were not prevented).

C.     Evaluated the IRS programming logic used to generate monthly Problem Correction Reports to identify accounts that need to have recertification indicators removed.  We also evaluated the request for computer programming to automatically remove recertification indicators for CY 2004 processing.  This programming would remove recertification indicators based on various Examination function closing actions.

D.    Discussed the required 5 percent review of specific nonexamined returns with IRS National Headquarters employees and 2 Compliance Sites.

II.                Evaluated the corrective action to Recommendation 3 in our December 2000 report.  Specifically, we evaluated whether the IRS procedures adequately advised IRS employees when to remove recertification indicators.

III.             Evaluated the corrective actions to Recommendations 4 and 5 in our December 2000 report regarding incorrectly suspended refunds.  Specifically, we determined whether the IRS identified and resolved incorrectly suspended refunds and whether there were systemic problems that would prevent suspended refunds from going to taxpayers. 

A.    From the IRS Individual Master File, obtained the identity of taxpayer accounts reflecting either of two specific freeze conditions that “suspended” refunds for EIC recertification returns as of August 4, 2003, and a file with the identity of the accounts included in the August 4, 2003, monthly Problem Correction Reports. 

B.     Performed various database queries and comparisons on these files to help evaluate the programming for the monthly Problem Correction Reports.  In addition, we determined whether the various IRS Compliance sites timely released refunds for accounts identified on the monthly Problem Correction Reports.  This included sending surveys to the Compliance sites as to whether the reports were worked.

IV.             Evaluated the corrective action to Recommendation 10 in our December 2000 report.  Specifically, we determined whether IRS communications with taxpayers clearly informed and educated taxpayers regarding the recertification process.

A.    Evaluated letters sent to taxpayers when the EIC is allowed for a recertification tax year to determine whether these letters specifically advise taxpayers that their EIC is being allowed and that the taxpayers are now recertified.

B.     Evaluated the letter (Computer Paragraph 79) sent to taxpayers when the EIC is denied during an examination.

C.     Evaluated the letter sent to taxpayers (math error letter for Math Error Code 653) when the IRS denies the EIC because taxpayers did not submit an Information to Claim Earned Income Credit After Disallowance (Form 8862) with their original paper tax returns.  We also determined whether Form 8862 and related Form 8862 Instructions were sent to taxpayers.

D.    Evaluated the electronically filed tax return reject message (Reject Code 600) wording used to inform taxpayers that their tax returns are being rejected because Forms 8862 were not filed.

V.                Evaluated the corrective action to Recommendation 11 in our December 2000 report.  Specifically, we determined whether the IRS advised potentially qualified taxpayers via letters (Computer Paragraph 27) that the taxpayers might be entitled to the income-only EIC; for example, taxpayers meeting age and income requirements and whose accounts have a recertification indicator value of 1.

VI.             Evaluated the corrective action to Recommendation 12 in our December 2000 report.  Specifically, we determined whether the IRS implemented our recommendation to consider changing the EIC Recertification Program regarding when taxpayers are recertified.  We evaluated the June 25, 2001, IRS regulations that adopted Recommendation 12 in our December 2000 report; discussed the regulations with personnel in the IRS National Headquarters and Office of Chief Counsel to determine the IRS’ intent regarding the June 25, 2001, regulations; and determined whether IRS programming for returns filed in CY 2003 correctly allowed the income-only EIC when no Form 8862 was filed.

 

Appendix II

 

Major Contributors to This Report

 

Michael R. Phillips, Assistant Inspector General for Audit (Wage and Investment Income Programs)

Scott A. Macfarlane, Director

Richard J. Calderon, Audit Manager

Larry A. Mart, Lead Auditor

Kathleen A. Hughes, Senior Auditor

John Kirschner, Senior Auditor

Arlene Feskanich, Information Technology Specialist

 

Appendix III

 

Report Distribution List

 

Commissioner  C

Office of the Commissioner – Attn:  Chief of Staff  C

Deputy Commissioner for Services and Enforcement  SE

Deputy Commissioner, Small Business/Self-Employed Division  SE:S

Deputy Commissioner, Wage and Investment Division  SE:W

Director, Communications and Liaison, Small Business/Self-Employed Division  SE:S:C&L

Director, Compliance, Wage and Investment Division  SE:W:CP

Director, Compliance Services Campus Operations, Small Business/Self-Employed Division  SE:S:CSCO

Director, Strategy and Finance, Wage and Investment Division  SE:W:S

Chief, Performance Improvement, Wage and Investment Division  SE:W:S:PI

Field Director, Compliance Services (Austin)  SE:W:CP:CS:AU

Field Director, Compliance Services (Fresno)  SE:W:CP:CS:F

Chief Counsel  CC

National Taxpayer Advocate  TA

Director, Office of Legislative Affairs  CL:LA

Director, Office of Program Evaluation and Risk Analysis  RAS:O

Office of Management Controls  OS:CFO:AR:M

Audit Liaisons: 

Chief, Customer Liaison, Small Business/Self-Employed Division  SE:S:COM

Senior Operations Advisor, Wage and Investment Division  SE:W:S

 

Appendix IV

 

Outcome Measures

 

This appendix presents detailed 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.

Type and Value of Outcome Measure:

·         Funds Put to Better Use – Potential; 14,800 taxpayers for $26.5 million (see page 5).

Methodology Used to Measure the Reported Benefit:

The $26.5 million for 14,800 taxpayers includes 3 different totals: 

1.  Total of $19.8 million from Samples 1 – 3 below.  This result affects 11,900 taxpayers.

Sample 1

 

4,900 taxpayers

 

$5.8 Million

Sample 2

 

4,800 taxpayers

 

$7.0 Million

Sample 3

 

2,200 taxpayers

 

$7.0 Million

Totals

 

11,900 taxpayers

 

$19.8 Million

 

2.  Total of $5.1 million from Sample 4.  This result affects 2,200 taxpayers.

3.  Total of $1.6 million from Item 5.  This result affects 700 taxpayers.

Sample 1

In July 2003, the Internal Revenue Service (IRS) removed approximately 25,000 recertification indicators from taxpayers’ accounts during a clean-up to remove indicators not previously removed when the Earned Income Credit (EIC) was allowed.  We took a random sample of
200 accounts and analyzed the accounts to determine whether the recertification indicators should have been removed.  This evaluation was only to determine whether the IRS allowed the EIC claimed during a recertification tax year.  We did not obtain the related tax returns to determine if the decisions to allow the EIC were correct. 
Results indicate that 39 (19.5 percent) of the 200 accounts had the indicators incorrectly removed. 

Ø      Using the 19.5 percent error rate times the population of 25,000 provides an estimate that 4,900 accounts incorrectly had the indicators removed.  This outcome results in a 95 percent confidence level with a +/- 6 percent precision level.

Ø      We estimate these 4,900 taxpayers received approximately $5.8 million (4,900 times the average of $1,180 per taxpayer) of EIC for Tax Year (TY) 2003 returns without the taxpayers having to demonstrate eligibility to receive it.  This estimate is +/- $2.5 million, for a range of $3.3 million to $8.3 million with a 90 percent confidence level.

The estimates will increase with each subsequently filed return on which these taxpayers claim the EIC.

Note:  Our estimates are conservative because our different samples did not take exception with those recertification indicators removed for “pick-up” tax year returns closed as no-change and on which the EIC was allowed.  Ideally, these would not meet recertification requirements.  However, not counting these taxpayers as being recertified would subject the taxpayers to a second examination to prove EIC eligibility.  We raised the issue to the IRS as to whether it would be appropriate for the IRS to make a business decision to consider these taxpayers to be recertified.

Sample 2

In addition to the July 2003 clean-up, IRS employees removed recertification indicators from approximately 66,000 accounts during the period January 2002 through August 4, 2003 (cycles 200201 through 200330).  We took a random sample of 250 accounts and analyzed the accounts to determine if the recertification indicators should have been removed.  This evaluation was only to determine if the IRS allowed the EIC claimed during a recertification tax year.  We did not obtain the related tax returns to determine if the decisions to allow the EIC were correct.

Sample results indicate that 18 (7.2 percent) of the 250 accounts had the indicators incorrectly removed.  Using the 7.2 percent error rate times the population of 66,000 provides an estimate that 4,800 accounts incorrectly had the indicators removed.  Based on these results, our projection has a 95 percent confidence level with a +/- 4 percent precision level.

Ø      We estimate these 4,800 taxpayers received over $7.0 million (4,800 times the average of $1,470 per taxpayer) of EIC for TYs 2002 and 2003 returns without the taxpayers having to demonstrate eligibility to receive the EIC.  This estimate is +/- $4.6 million, for a range of $2.4 million to $11.6 million with a 90 percent confidence level.

Ø      We believe these estimates are conservative because our only criterion for determining whether the recertification indicators were correctly removed for this sample was whether the EIC claimed was allowed during a recertification tax year.  Additionally, the estimates will increase with each subsequently filed return on which these taxpayers claim the EIC.

Sample 3

We identified approximately 12,000 tax returns closed as nonexamined and reflecting an EIC Recertification Program Examination Project Code for the period August 12, 2002, through August 4, 2003 (cycles 200231 through 200330).  We excluded various categories of nonexamined closures (returns claiming the income-only EIC, accounts with Criminal Investigation function transactions, returns intended by the IRS National Headquarters to be closed as nonexamined via filter programming, and other miscellaneous categories).  We excluded these categories to provide a population for which Examination function employees would have decided to allow the child-related EIC. 

This left 4,900 returns from which we took a random sample of 200 and determined if the decisions to close the returns as nonexamined were appropriate.  Closing the returns as nonexamined would recertify the taxpayers, and the recertification indicators should have been removed.  When necessary, we obtained the related tax returns to help determine if the decisions to allow the EIC were correct.

Sample results indicate 90 (45 percent) of the 200 returns had incorrect decisions to allow the EIC, which resulted in an average of $3,200 in EIC per taxpayer being allowed.

Ø      Using the 45 percent error rate times the population of 4,900 provides an estimate that 2,200 accounts had the EIC incorrectly allowed.  Based on these results, our projection has a 95 percent confidence level with a +/- 7 percent precision level.

Ø      We estimate these 2,200 taxpayers received over $7.0 million of EIC (2,200 taxpayers times the average of $3,200 per taxpayer) for 1 or more tax periods without the taxpayers having to demonstrate eligibility to receive the EIC.  This estimate is +/- $1.3 million, for a range of $5.7 million to $8.3 million with a 90 percent confidence level.

We believe the $7.0 million amount is conservative.  For example, if we had used a 19-month sample period rather than a 12-month sample period, we would have had a population of 11,600 closures rather than 4,900 and our monetary projection would be higher.

Sample 4

For the period January 2003 through August 4, 2003, we obtained from the IRS Individual Master File the identity of approximately 8,000 taxpayer accounts for which “filter programming” prevented examinations.  This included accounts reflecting (1) an EIC recertification indicator value of “1,” (2) an Audit [Examination] Code “U” indicating an Information to Claim Earned Income Credit After Disallowance (Form 8862) had been filed, (3) child-related EIC claimed, and (4) no Examination function-related transaction codes.  We took a random sample of 200 of these 8,000 accounts and reviewed the accounts to determine if the programming correctly prevented examinations when the EIC had been previously allowed without removing the recertification indicators.

Sample results indicate that 56 (28 percent) of the 200 accounts had the EIC incorrectly allowed without taxpayers having to demonstrate that they were eligible to receive the EIC, which resulted in an average of $2,300 in EIC per taxpayer being allowed.

Ø      Using this 28 percent error rate times the population of 8,000 provides an estimate that 2,200 accounts incorrectly had the EIC allowed.  Based on these results, our projection has a 95 percent confidence level with a +/- 7 percent precision level.

Ø      We estimate these 2,200 taxpayers received approximately $5.1 million in EIC (2,200 taxpayers times the $2,300 average per taxpayer) for TY 2002 returns without the taxpayers having to demonstrate eligibility to receive the EIC.  This estimate is +/- $1.1 million, for a range of $4.0 million to $6.2 million with a 90 percent confidence level.

Item 5

In February 2003, the IRS filter programming identified approximately 6,100 returns processed in January and February 2003 for which examinations would not be conducted.  We evaluated the filter programming used to identify these returns to determine if the programming prevented examinations (filtered out returns) for only those taxpayers that should be considered recertified.

Based on our findings, the IRS re-ran the filter programming for the 6,100 returns and identified approximately 700 returns for which examinations should take place.  The EIC claimed for these returns is estimated to be approximately $1.6 million.  This estimate is based on an average EIC allowed of approximately $2,340 for approximately 3,700 filter returns, closed as nonexamined, on which the EIC amounts were allowed.

We believe the $1.6 million amount is conservative because changes to IRS programming, based on our findings, would have helped safeguard revenue for returns processed after February 2003.

Type and Value of Outcome Measure:

·         Taxpayer Rights and Entitlements – Actual; 26,300 taxpayers for $9 million (see pages 11 and 18).

Methodology Used to Measure the Reported Benefit:

The $9.0 million for 26,300 taxpayers includes 3 different totals:

  1. Total of $3.9 million from Item 1.  This result affects 3,200 taxpayers.
  2. Total of over $400,000 from Item 2.  This result affects over 2,100 taxpayers.
  3. Total of $4.7 million from Item 3.  This result affects over 21,000 taxpayers.

Item 1 – Delayed Refunds

We obtained from the IRS Individual Master File a file containing the identity of taxpayer accounts reflecting either of two specific freeze conditions (Transaction Codes 570 and 810) that “suspended” refunds for EIC recertification as of August 4, 2003 (cycle 200330).  We then obtained from the IRS a file with the identity of the accounts included in the August 4, 2003, monthly Problem Correction Reports.  We then matched the two files to determine if all “suspended” refund accounts were properly included in the IRS Problem Correction Accounts.  We also surveyed each Compliance site as to whether they resolved the “suspended” refunds reported to them on the monthly Reports.  Our file match and surveys identified the following situations involving accounts with delayed refunds:

Ø      Approximately 900 accounts, with suspended refunds totaling approximately $1.6 million, which did not appear on the monthly Problem Correction Reports.

Ø      Approximately 1,900 accounts with suspended refunds totaling approximately $1.8 million.  The monthly Problem Correction Reports were not worked at all for four Compliance sites.

Ø      Approximately 400 accounts with suspended refunds totaling approximately $500,000.  The monthly Problem Correction Reports were not worked timely at 2 Compliance sites (they did not work the reports within 3 weeks of receiving the reports).

The $3.9 million and 3,200 taxpayer figures are conservative.  For example:

Ø      There are five different Problem Correction Reports that come out monthly.  By definition, each account that appeared on some of these Reports had delayed refunds, but our outcome includes only those accounts not worked or not worked timely.

Ø      We computed the delay only from the time the Examination function closed the case (as evidenced by the Transaction Code 421 posting cycle), rather than using the time the refund was actually suspended when the related tax returns initially posted.

Ø      We computed our delay to only 3 weeks after cycle 200330 (our extract cycle) and not to the actual cycle the refunds were issued, which could have been as much as 1 year or more after our referral to the IRS (this occurred for approximately 1,000 of the 3,200 refunds).

Ø      Our totals would not have included all the accounts over a several year period for which taxpayers would have contacted the IRS to receive their delayed refunds.

Item 2 – Incorrectly Denied Income-Only EIC

This included 2,100 taxpayers and $400,000, which are totals for the following 2 situations:

Ø      In February 2003, we identified a programming problem with not allowing the income-only EIC for TY 2002.  The IRS subsequently did a clean-up that identified 900 taxpayers who were then allowed the income-only EIC.

Ø      In August 2003, based on our IRS Individual Master File extracts, we identified another 1,200 taxpayers who did not receive the income-only EIC and referred these to the IRS for corrective action.  This included returns processed during a 2-week period in March 2003 that the IRS did not identify during the clean-up mentioned immediately above.  However, as of November 2004 and subsequent to the end of fieldwork, the IRS had not taken corrective action to ensure these taxpayers received the income-only EIC.

Note:  The IRS also denied income-only EIC totaling over $100,000 for an additional 400 taxpayers, but these taxpayers contacted the IRS and were allowed the EIC prior to our August 2003 referral.  The fact that only 400 taxpayers questioned the EIC denial indicates taxpayers are often confused to the point they will not protest having their EIC incorrectly denied.

Item 3 – Changes to Computer Programming

The IRS took corrective action to computer programming problems that denied income-only EIC for TY 2002 returns.  This then allowed correct processing for over 21,000 taxpayers to receive income-only EIC amounting to over $4.7 million in 2003.  We identified these taxpayers from queries of information obtained from the IRS Individual Master File for tax returns posting from January 2003 through August 4, 2003 (cycles 200301 through 200330).  This included tax returns claiming income-only EIC for accounts with recertification indicators having value 1.

Type and Value of Outcome Measure:

·         Taxpayer Rights and Entitlements – Potential; 10,500 taxpayers for $20.8 million (see page 5).

Methodology Used to Measure the Reported Benefit:

We obtained the identity of taxpayer accounts reflecting recertification indicator values of 1 from a 1 percent sample of all taxpayers on the IRS Individual Master File as of April 7, 2003.  This sample included 8,760 taxpayer accounts and projects to a population of 876,000 accounts having the indicator as of April 7, 2003.  We took a random sample of 250 accounts and analyzed the accounts to determine if the recertification indicators were appropriate.  If the indicators had not been removed when the EIC claims were allowed, we determined if subsequent EIC claims were denied.

Sample results indicate 15 (6 percent) of the 250 accounts did not have the indicators removed when the EIC claimed was allowed.  Three (1.2 percent of the 250) of these 15 accounts later had EIC denied for an average EIC dollar amount of $4,100 per account.  Projecting these error rates to the population of 876,000 accounts, we determined there were approximately 52,000 accounts that still had an erroneous indicator.  Of these 52,000 accounts, 10,500 had the EIC later denied improperly (see next 2 bullets), and 17,000 had erroneous examinations performed (see last section).

Ø      Using the 1.2 percent error rate times the population of 876,000 provides an estimate that 10,500 accounts had the EIC later denied.  Based on these results, our projection has a 95 percent confidence level with a +/- 2 percent precision level.

Ø      We estimate these 10,500 accounts then had the EIC later denied for an estimated $20.8 million.  We arrived at this dollar amount by multiplying 10,500 times $1,980, which is the average amount of child-related EIC allowed by the IRS for tax returns processed during the Calendar Years (CY) 2001 to 2003.

We used this logic for computing the dollar amount to provide a conservative figure as compared to the significant range that would otherwise be obtained using statistical projections.  The final figure of $20.8 million would be in the lower 25 percent of that dollar range.

The $20.8 million is also conservative because our sample was taken at a “point in time” as of April 2003.

Type and Value of Outcome Measure:

·         Taxpayer Burden – Potential; approximately 480,000 letters or electronic messages sent to taxpayers that did not clearly or accurately communicate the EIC recertification requirements, and over 50,000 taxpayers who filed unnecessary tax forms (see page 13).

Methodology Used to Measure the Reported Benefit:

See the first 3 items below that comprise the 480,000.  The 370,000 Computer Paragraph (CP) 79 letters mentioned in Item 4 below are not being claimed as an outcome for this report because a May 2004 Treasury Inspector General for Tax Administration report has already claimed an outcome measure (for problems different from those we currently identified).

Item 1 – Not Specifically Advising Taxpayers They Were Recertified

The IRS did not specifically advise over 100,000 taxpayers they were recertified when the IRS removed recertification indicators during CYs 2002 and 2003.  The over 100,000 figure would be the total of (1) 66,000 recertification indicators removed by individual IRS employees based on our computer analysis of transactions posting to the IRS Individual Master File for the period January 2002 through August 4, 2003; (2) 25,000 removed via a July 2003 IRS clean-up; (3) 9,600 removed in December 2003 by additional computer programming as partial corrective action to problems we referred to the IRS in 2003; and (4) other recertification indicators (we did not obtain the actual number) removed by individual IRS employees from August 5, 2003, through December 31, 2003.

Item 2 – Forms 8862 Not Included With Math Error Letters

The IRS sent over 200,000 Math Error 653 letters total during CYs 2001 through 2003.  These letters stated that Form 8862 was being included with the letter, but the Form 8862 was not included.  The IRS provided us the volumes of these letters sent that included approximately 73,800; 93,100; and 35,700 for CYs 2001, 2002, and 2003, respectively.

Item 3 – Inadequate Explanations for Denying the EIC

Both electronically filed tax returns and paper tax returns did not adequately explain why the EIC was being denied for over 210,000 taxpayers for CY 2003 processing.

Ø      If taxpayers subject to recertification do not file a Form 8862 with their electronically filed tax returns, the IRS rejects the returns and provides a reject message.  For CY 2003, the IRS rejected over 176,000 electronic returns for not having a Form 8862.  We obtained this information from the weekly Service Center Error Reject Codes Report dated October 3, 2003, for Reject Code 600.

Ø      If taxpayers do not file a Form 8862 with their paper returns, the IRS denies the EIC and sends a letter to the taxpayers.  For CY 2003, this would include the same 35,700 returns discussed above for Math Error 653 letters not having Forms 8862 for CY 2003, so these 35,700 are not included in the overall total of approximately 480,000 communications with problems.

These numbers are very conservative because they do not include volumes prior and subsequent to CY 2003 processing.

Item 4 – Problems With Letter CP 79

Corrective action to letter CP 79 resulted in inaccurate information and clarity issues.  The IRS sent over 370,000 of these letters during 2003.

Item 5 – Form 8862 Not Needed and/or Incorrectly Requiring Form 8862

During CY 2003, over 50,000 taxpayers filed a Form 8862 that was not needed when filing an electronic or paper return.  This resulted in part because (1) the IRS did not specifically advise taxpayers they were recertified or (2) the IRS had programming problems that incorrectly required taxpayers to file a Form 8862 with electronic TY 2002 returns.

Ø      The IRS electronic filing (ELF) weekly processing reports track the number of tax returns for which a Form 8862 is not needed at the times the ELF returns are processed.  Specifically, these are returns with a Form 8862 and a recertification indicator value of zero (0) per the report heading “8862 EITC Acc: No Recertification/0.”  This included over 30,000 returns processed in CY 2003.

Ø      From the IRS Individual Master File, we identified tax returns posting from January 2003 through August 4, 2003, (cycles 200301 through 200330) with an Audit [Examination] Code U (indicating Form 8862 is present).  Our queries identified, for TY 2002, approximately 15,000 electronic returns and approximately 5,000 paper returns claiming the EIC that would not require a Form 8862.

We estimate that completing the Forms 8862 could have wasted over 50,000 hours of taxpayer time.  The 50,000 hours is intended to be conservative and was computed by multiplying the number of unnecessary Forms 8862 (50,000) by 1 hour for each Form.  The Form 8862 Instructions stated that it would take about 2 hours and 20 minutes of time to complete the Form 8862.  However, many of the 50,000 returns would relate to income-only EIC and the Form 8862 is easier to complete for these taxpayers.

Type and Value of Outcome Measure:

·         Taxpayer Burden – Potential; 17,000 taxpayers (see page 5).

Methodology Used to Measure the Reported Benefit:

The methodology for this outcome uses the same random sample of 250 accounts used for the outcome previously discussed for “Taxpayer Rights and Entitlements – Potential; 10,500 taxpayers for $20.8 million.”

Sample results indicate 15 (6 percent) of the 250 accounts did not have the indicators removed when the EIC claimed was allowed.  When these sample results are projected to the population of 876,000 accounts, we found 52,000 accounts still had erroneous indicators.  Five (2.0 percent of the 250) of these 15 accounts later had tax returns examined because the recertification indicators had not been removed.

Using this 2.0 percent error rate times the population of 876,000 provides an estimate that of the 52,000 accounts with erroneous indicators, approximately 17,000 taxpayers incorrectly had their tax returns examined.  Based on these results, our projection has a 95 percent confidence level with a +/- 2 percent precision level.

 

Appendix V

 

Management’s Response to the Draft Report

 

The 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.