GAO-07-1110, Advance Earned Income Tax Credit: Low Use and Small Dollars Paid Impede IRS's Efforts to Reduce High Noncompliance


This is the accessible text file for GAO report number GAO-07- entitled 
'Advance Earned Income Tax Credit: Low Use and Small Dollars Paid 
Impede IRS's Efforts to Reduce High Noncompliance' which was released 
on August 31, 2007. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as part 
of a longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the printed 
version. The portable document format (PDF) file is an exact electronic 
replica of the printed version. We welcome your feedback. Please E-mail 
your comments regarding the contents or accessibility features of this 
document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

Report to the Joint Committee on Taxation: 

United States Government Accountability Office: 

GAO: 

August 2007: 

Advance Earned Income Tax Credit: 

Low Use and Small Dollars Paid Impede IRS's Efforts to Reduce High 
Noncompliance: 

Advance Earned Income Tax Credit: 

GAO-07-1110: 

GAO Highlights: 

Highlights of GAO-07-1110, a report to the Chairman and Vice Chairman 
of the Joint Committee on Taxation. 

Why GAO Did This Study: 

The Advance Earned Income Tax Credit (AEITC) allows individuals to 
receive a portion of the Earned Income Tax Credit (EITC) in their 
paychecks, instead of receiving all of it when filing their year-end 
tax return. Limited research has been conducted on the AEITC since GAO 
last examined it in the early 1990s. 

GAO was asked to determine: 
(1) how many individuals received the AEITC compared with the EITC in 
tax years 2002 through 2004, what actions, if any, have been taken to 
increase use, and the potential for increases in use in the future; (2) 
the extent of noncompliance, if any, associated with the AEITC; and (3) 
how well the Internal Revenue Service’s (IRS) procedures address the 
areas of noncompliance. To address these questions, GAO analyzed Forms 
W-2 and tax return data and interviewed IRS and Social Security 
Administration (SSA) officials 

What GAO Found: 

AEITC use was low—only about 3 percent of EITC recipients potentially 
eligible for the advance received it in tax years 2002 through 2004, or 
about 514,000 of the 17 million potentially eligible individuals each 
year. About half of all recipients received $100 or less in AEITC and 
75 percent received $500 or less for the year, with a total benefit 
paid of about $146 million each year. Several efforts have been aimed 
at increasing use over the last approximately15 years, such as sending 
notices to individuals informing them that they were potentially 
eligible for the AEITC and making changes to IRS forms. Despite these 
efforts, use did not substantially increase and, for several reasons, 
it may be difficult to increase it in the future. For example, IRS 
officials, other experts, and prior GAO work suggests that individuals 
often do not elect the AEITC because they prefer receiving the entire 
EITC as a lump sum after filing their tax return. 

As many as 80 percent of AEITC recipients did not comply with at least 
one of the program requirements GAO reviewed, and some were 
noncompliant with more than one during the 3 years we reviewed. In tax 
years 2002 through 2004, about 20 percent, or more than 100,000 AEITC 
recipients, may not have been eligible for the AEITC because they had 
an invalid Social Security number (SSN). These individuals received a 
total of $37 million to $39 million each year. Almost 40 percent (about 
200,000 recipients) did not file the required tax return; these 
individuals received $42 million to $50 million each year. Of the about 
60 percent (more than 300,000) AEITC recipients who did file a return, 
about two-thirds misreported the amount received. 

Figure: AEITC Use, Dollars and Compliance In Tax Years 2002-2004. 

[See PDF for image] 

Source: GAO analysis of IRS data. 

[End of figure] 

IRS’s procedures have limited effectiveness in addressing AEITC 
noncompliance. For example, Automated Underreporter (AUR) staff worked 
on only a fraction of AEITC cases because of resource constraints and 
criteria limiting case selection. IRS could address AEITC noncompliance 
by sending “soft notices” to recipients, requiring employers to verify 
employee SSNs before providing the AEITC, or creating a Forms W-5, 
“EITC Advance Payment Certificate,” database. Each of these options 
have advantages, however, they also have potential disadvantages that 
could limit their effectiveness. 

What GAO Recommends: 

GAO recommends that the Acting Commissioner of Internal Revenue analyze 
options to reduce AEITC noncompliance such as implementing a “soft 
notice” test. If these options are deemed ineffective, and no other 
options are viable, the Treasury Secretary should provide an opinion to 
the Congress as to whether the AEITC should be retained. IRS agreed 
with our recommendation and outlined the actions IRS plans to take. 

[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-07-1110]. 

To view the full product, including the scope and methodology, click on 
the link above. For more information, contact Michael Brostek at (202) 
512-9110 or brostekm@gao.gov. 

[End of section] 

Contents: 

Letter: 

Results in Brief: 

Background: 

AEITC Use Low and May Be Difficult to Increase: 

High Noncompliance Exists with AEITC Requirements: 

IRS's AEITC Compliance Procedures Have Limited Effectiveness and 
Options to Better Address Noncompliance Have Advantages and 
Disadvantages: 

Conclusions: 

Recommendations for Executive Action: 

Agency Comments: 

Appendix II: RS's Implementation of Recommendations from GAO's 1992 
Report on the Advance Earned Income Tax Credit: 

Appendix II: Scope and Methodology: 

Appendix III: Demographic Characteristics of Advance Earned Income Tax 
Credit Recipients and Their Employers: 

Appendix IV: Administrative and Legislative Changes to the Advance 
Earned Income Tax Credit Since 1990: 

Appendix V: Additional Analyses of Advance Earned Income Tax Credit 
Noncompliance: 

Appendix VI: Comments from the Internal Revenue Service: 

Appendix VII: GAO Contacts and Staff Acknowledgments: 

Tables: 

Table 1: EITC Requirements for 2006 That All AEITC Recipients Must Also 
Expect to Meet: 

Table 2: Percentage of AEITC Recipients Receiving an Amount of AEITC, 
Tax Years 2002 through 2004: 

Table 3: Frequency of Individuals Electing the EITC and AEITC, Tax 
Years 2002 through 2004: 

Table 4: AEITC Recipients and Recipients with Invalid SSN on Form W-2, 
Individuals and Dollars, Tax Years 2002 through 2004: 

Table 5: AEITC Recipients and Recipients Who Did Not File a Tax Return, 
Individuals and Dollars, Tax Years 2002 through 2004: 

Table 6: AEITC Recipients, Recipients Who Did Not File a Federal Tax 
Return, and Recipients Who Did Not File a Tax Return and Had an Invalid 
SSN on the Form W-2, Individuals and Dollars, Tax Years 2002 through 
2004: 

Table 7: AEITC Matches and Mismatches between Form(s) W-2 and Filed Tax 
Returns, Tax Years 2002 through 2004: 

Table 8: IRS Implementation of Recommendations from GAO's 1992 Report 
on the Advance Earned Income Tax Credit: 

Table 9: Number of Forms W-2, Number of Individuals, and the Dollars 
for Each Subpopulation, Tax Years 2002 through 2004: 

Table 10: Number of Forms W-2 Received by Subpopulation, Tax Years 2002 
through 2004: 

Table 11: AGI for AEITC and EITC Recipients, Number, Average, and Sum, 
for Tax Year 2002: 

Table 12: AGI for AEITC and EITC Recipients, Number, Average, and Sum, 
for Tax Year 2003: 

Table 13: AGI for AEITC and EITC Recipients, Number, Average, and Sum, 
for Tax Year 2004: 

Table 14: Wages Reported for AEITC and EITC Recipients, Number, Wage 
Amount, and Average Dollars Received for Tax Years 2002 through 2004: 

Table 15: Filing Status of AEITC and EITC Recipients, Number and 
Percentage, Tax Years 2002 through 2004: 

Table 16: Age of AEITC and EITC Recipients, Tax Years 2002 through 
2004: 

Table 17: Gender of AEITC and EITC Recipients, Number and Percentage, 
Tax Years 2002 through 2004: 

Table 18: Number of Qualifying Children for AEITC and EITC Recipients, 
Number and Percentage, Tax Years 2002 through 2004: 

Table 19: Filing Method of AEITC and EITC Recipients, Number and 
Percentage, Tax Years 2002 through 2004: 

Table 20: Employer Size, Number of Employers, Number of Forms W-2 
Employers Issued to AEITC Recipients and Total Dollars Reported on 
Forms W-2, Tax Years 2002 through 2004: 

Table 21: Administrative Changes to the AEITC since 1990: 

Table 22: Number of Individuals and AEITC Dollars Received by AEITC 
Recipients and Individuals Who Elected the AEITC Consecutively Using an 
Invalid SSN and an Invalid SSN and Not Filing, Tax Years 2002 through 
2004: 

Table 23: Number of AEITC Consecutive Recipients with Matches and 
Mismatches between Forms W-2 and Filed Tax Returns, Tax Years 2002 
through 2004: 

Table 24: Amount of AEITC Received Per Invalid Form W-2, Number and 
Percentage, Tax Years 2002 through 2004: 

Table 25: Number of Individuals Receiving a Form(s) W-2, Total Number 
of Form(s) W-2 Reporting AEITC, Total Associated and Average Dollars 
Received Relative to Number of Form(s) W-2 Reporting AEITC for AEITC 
Recipients with an Invalid SSN on the Form(s) W-2, Tax Year 2002: 

Table 26: Number of Individuals Receiving a Form(s) W-2, Total Number 
of Form(s) W-2 Reporting AEITC, Total Associated and Average Dollars 
Received, Relative to Number of Form(s) W-2 Reporting AEITC for AEITC 
Recipients with an Invalid SSN on the Form(s) W-2, Tax Year 2003: 

Table 27: Number of Individuals Receiving a Form(s) W-2, Total Number 
of Form(s) W-2 Reporting AEITC, Total Associated and Average Dollars 
Received, Relative to Number of Form(s) W-2 Reporting AEITC for AEITC 
Recipients with an Invalid SSN on the Form(s) W-2, Tax Year 2004: 

Figures: 

Figure 1: AEITC Process for Employees and Employers: 

Figure 2: Number of EITC Recipients, EITC Recipients Potentially 
Eligible for the AEITC, AEITC Recipients, and Total Dollars They 
Received, Tax Years 2002 through 2004: 

Figure 3: Percentage of AEITC Recipients Compliant and Noncompliant 
with at Least One AEITC Requirement, Tax Years 2002 through 2004: 

Figure 4: Submission Processing Procedures for Paper and Electronic Tax 
Returns Reporting AEITC: 

Figure 5: AUR Program Procedures to Resolve Discrepancies between AEITC 
Amounts on Forms W-2 and Tax Returns: 

Figure 6: Average Amount of AEITC Received by Consecutive and AEITC 
Recipients, Tax Years 2002 through 2004: 

Figure 7: Geographic Location of AEITC Recipients in the Valid 
Subpopulation, Tax Years 2002 through 2004: 

Figure 8: Geographic Location of AEITC Recipients in the Invalid Name 
Subpopulation, Tax Years 2002 through 2004: 

Figure 9: Geographic Location of AEITC Recipients in the Invalid Number 
Subpopulation, Tax Years 2002 through 2004: 

Figure 10: Number of Forms W-2 Received with Amounts above the Yearly 
AEITC Maximum, Tax Years 2002 through 2004: 

Figure 11: AEITC Dollars Reported on Form W-2s above the Yearly AEITC 
Maximum, Tax Years 2002-2004: 

Abbreviations: 

AEITC: Advance Earned Income Tax Credit AGI Adjusted Gross Income: 
AUR: Automated Underreporter: 
DM-1: Data Master File: 
EITC: Earned Income Tax Credit: 
ERS: Error Resolution System: 
IRS: Internal Revenue Service: 
OBRA: '93Omnibus Budget Reconciliation Act of 1993: 
SSA: Social Security Administration: 
SSN: Social Security number: 
SSNVS: Social Security Number Verification System: 
TIN: Taxpayer Identification Number: 

United States Government Accountability Office: 

Washington, DC 20548: 

August 10, 2007: 

The Honorable Charles B. Rangel: 
Chairman: 
The Honorable Max Baucus: 
Vice Chairman: 
Joint Committee on Taxation: 

Beginning in 1979, eligible taxpayers could elect to receive the Earned 
Income Tax Credit (EITC), a refundable credit available to low income 
workers, as an advance payment--an option currently found in only one 
other federal tax credit.[Footnote 1] Individuals who elect this option 
receive a portion of the EITC from their employer throughout the year 
with their regular pay, instead of receiving only a lump sum refund or 
tax credit when filing their year end federal income tax 
returns.[Footnote 2] The Advance Earned Income Tax Credit (AEITC) 
provides employees an immediate benefit from the credit, which may help 
them meet their daily expenses by increasing their take-home pay. 

Little has been known about the individuals who receive the AEITC, the 
employers who pay it, and compliance with its requirements. Neither the 
Internal Revenue Service (IRS) nor the employer is required to confirm 
the eligibility of those who elect the AEITC before they receive it. 
Thus, the extent of compliance problems related to the AEITC may depend 
on factors such as how knowledgeable individuals applying for the AEITC 
are about eligibility requirements and how well the IRS addresses 
various kinds of noncompliance that the agency may identify only after 
AEITC payments have been made. 

Limited research has been conducted on the AEITC since we last examined 
it in the early 1990s. In 1992, we reported that the percentage of 
individuals receiving the EITC in 1989 who also received the AEITC was 
low and that many eligible workers were not aware of it.[Footnote 3] We 
estimated that almost half of those who received the AEITC in 1989 and 
filed a tax return did not report receiving it and that 45 percent of 
people whom IRS records show may have received the AEITC never filed a 
tax return, meaning that IRS would have no way of determining their 
eligibility. We made six recommendations to the Commissioner of 
Internal Revenue designed to increase awareness of the AEITC among 
eligible individuals and improve compliance among those who receive it. 
Appendix I discusses IRS's implementation of the recommendations from 
our 1992 AEITC report. 

To provide current information on the AEITC, the Committee asked us to 
answer the following questions: (1) How many individuals received the 
AEITC compared with the EITC and how much did they receive in tax years 
2002 through 2004? What actions, if any, have been taken to increase 
use since 1992 and what is the potential for significant increases in 
the future? (2) What is the extent of noncompliance, if any, associated 
with the AEITC? (3) How well do IRS's procedures address any areas of 
noncompliance? In addition, the Committee asked us to provide basic 
demographic characteristics of AEITC recipients and their employers and 
to identify legislative and administrative changes made since our 1992 
report, some of which were discussed by IRS in its response to that 
report's recommendations. The information is presented in appendixes I 
and III through V. 

To answer these questions, we obtained a data file from IRS containing 
all Form W-2 "Wage and Tax Statements" that reported AEITC in tax years 
1999 through 2004. We performed data reliability tests on this file to 
determine whether the data were sufficiently reliable for our intended 
purposes and determined that they were. This process involved, among 
other things, checking the validity of the Social Security number (SSN) 
on the Form W-2, determining whether the AEITC amount was above the 
yearly maximum limit, and creating subpopulations in which to conduct 
our analyses. Using the Form W-2 data and data from other sources, such 
as IRS's Individual Return Transaction File that contains tax return 
information, we compiled relevant demographic characteristics data and 
compared AEITC data with EITC data when possible. We also reviewed 
administrative and legislative changes as well as relevant reports and 
considered these along with our data analyses. To examine actions 
intended to increase use and determine the potential for increases in 
future AEITC use, we reviewed various reports and literature and 
interviewed IRS officials and individuals we determined to be experts. 
We also used the Form W-2 and other data to identify areas of 
noncompliance by comparing the data to the AEITC requirements. In 
addition, we examined IRS's procedures for processing individual income 
tax returns with AEITC and interviewed IRS officials to determine how 
IRS addressed noncompliance. Finally, we explored possible remedies for 
the noncompliance we identified through literature searches and 
interviews with various Department of the Treasury, IRS, and Social 
Security Administration (SSA) officials. It was not within the scope of 
our work to fully evaluate the potential cost and benefits of these 
options for reducing noncompliance. We conducted our work in accordance 
with generally accepted government auditing standards from December 
2005 through July 2007. Appendix II provides more detail on the scope 
and methodology we used to conduct our work. 

Results in Brief: 

AEITC use and the amount paid out by employers was low in tax years 
2002 through 2004--only about 3 percent of the EITC recipients 
potentially eligible for the advance received it and about half of the 
recipients received $100 or less per year. In total, approximately 
514,000 individuals received about $146 million of AEITC in each of the 
3 years. Several federal efforts, including two legislative and several 
administrative changes, have been aimed at increasing AEITC use in the 
last approximately 15 years. For example, the Omnibus Budget 
Reconciliation Act of 1993 (OBRA '93), directed IRS to send notices to 
taxpayers during 2 tax years informing them that they were potentially 
eligible for the advance. Similarly, IRS made changes to its forms and 
increased speaker seminars to inform the public about the AEITC. 
Despite these efforts, use has not substantially increased and, for 
several reasons, it may be difficult to significantly increase it in 
the future. For example, IRS officials, other experts, and our prior 
work suggest that individuals often do not elect the AEITC because they 
prefer receiving the entire EITC as a lump sum when filing their tax 
return. An IRS-funded study revealed that low use is also attributed to 
concerns that prospective recipients would receive more in advance than 
they would ultimately be entitled to, thus owing the difference when 
filing their tax return. 

As many as 80 percent of recipients did not comply with at least one 
AEITC requirement, while some individuals were noncompliant with more 
than one requirement. Approximately 20 percent (more than 100,000) of 
all AEITC recipients each year may not have been eligible for the 
advance because they did not have a valid SSN, that is, a SSN that has 
been assigned by the SSA or a SSN/name combination that could be 
matched to SSA records. These individuals received $37 million to $39 
million in each tax year from 2002 through 2004. Almost 40 percent 
(about 200,000 recipients) did not file the required tax return; they 
received between $42 million and $50 million of AEITC for each of these 
years. When individuals do not file a return, IRS cannot readily 
identify whether they were eligible for AEITC and whether they owed IRS 
any of the amounts they received. Conversely, by not filing a tax 
return, some individuals did not receive additional EITC monies that 
they could only receive had they filed. Of the 60 percent of AEITC 
recipients who did file a return, about 300,000 individuals in each of 
the 3 years, two-thirds misreported the amount they received--the 
majority did not report receiving any AEITC. As a result, these AEITC 
recipients who claimed the EITC when filing their return collectively 
received an excess amount of EITC of about $64 million for the 3 years. 

Although IRS has processing and enforcement procedures in place to 
monitor the AEITC, the agency has limited effectiveness in addressing 
AEITC noncompliance, in large part because it targets most of its 
enforcement efforts at more widely used programs that involve larger 
amounts of money. For example, Automated Underreporter (AUR) staff 
worked only a fraction of the AEITC underreporting cases because 
resource constraints and criteria limited case selection. Moreover, AUR 
and other IRS enforcement programs use revenue collection potential as 
a primary criterion in identifying cases on which to work, and AEITC 
cases generally have low revenue potential. Although all AEITC nonfiler 
cases are eligible to be worked on by various enforcement programs, IRS 
officials said they work on only a limited number for similar reasons. 

One means to improve AEITC compliance might be to send "soft notices." 
Recent IRS tests have shown that the agency has successfully reduced 
subsequent noncompliance in situations that involve relatively small 
amounts of money by sending soft notices that ask taxpayers to 
voluntarily fix their misreporting by filing an amended return or not 
repeating the action in the next year. However, high annual turnover in 
the AEITC population could undermine this approach. Another means to 
improve AEITC compliance would be to partially verify individuals' 
eligibility before they receive the advance by verifying their SSNs. 
IRS operates the Taxpayer Identification Number (TIN) Matching service 
and SSA maintains the Social Security Number Verification System 
(SSNVS), both of which are open to registered public users for the 
purpose of determining the validity of an SSN. Both have unique 
advantages and disadvantages that would come into play if employers 
were required to use them to verify the SSNs of individuals seeking 
AEITC. SSA emphasized that verifying eligibility for the AEITC is most 
appropriate for IRS because it is a tax administration issue and 
therefore outside the scope of SSA's mission. Separately, IRS could 
verify individuals' SSNs and take other enforcement steps if it had a 
database of AEITC applicants. However, IRS cannot create such as 
database because it does not require employers to send copies of the 
Form W-5, "Earned Income Credit Advance Payment Certificate," to them 
when an employee requests receipt of the AEITC. Despite some benefits, 
IRS officials questioned the effectiveness of such a requirement due to 
(1) the low anticipated return on investment of AEITC compared with 
other areas with noncompliance problems and (2) their experience with 
the prior Questionable W-4 program, which required employers to submit 
information to IRS on taxpayers claiming more than 10 withholding 
allowances or exemptions, that showed employers did not adhere to 
similar requirements imposed by the agency. 

We recommend that the Acting Commissioner of Internal Revenue analyze 
whether it could cost effectively and significantly reduce AEITC 
noncompliance by sending soft notices to potentially noncompliant AEITC 
recipients, requiring employers to verify the SSN of employees seeking 
the AEITC, or requiring employers to submit Forms W-5 to IRS to create 
an AEITC database. To better identify the costs and implementation 
issues as well as the likelihood for these or other options to reduce 
AEITC noncompliance, where practical, the Acting Commissioner of 
Internal Revenue should test these options to make a more fully 
informed judgment about whether any would be worthwhile. If the Acting 
Commissioner of Internal Revenue determines that none of these options 
would be cost effective and that no other remedies are viable, then the 
Treasury Secretary should inform the Congress of this and provide 
Treasury's opinion about whether the AEITC should be retained. 

In providing written comments on a draft of this report (see app. VI), 
the Acting Commissioner of Internal Revenue agreed with our 
recommendation and outlined the actions IRS would take to address that 
recommendation, including conducting further analyses and possible 
testing of proposed options for reducing AEITC noncompliance. He also 
stated that IRS will conduct its cost-benefit analyses in conjunction 
with a congressional requirement to study the impact of expanding 
eligibility of the AEITC to all EITC recipients.[Footnote 4] We also 
provided a draft of this report to the Department of the Treasury and 
SSA and incorporated technical comments where appropriate. SSA 
emphasized that verifying eligibility for the AEITC is most appropriate 
for IRS because it is a tax administration issue and outside the scope 
of SSA's mission. 

Background: 

The EITC was enacted in 1975 and was originally intended to offset the 
burden of Social Security taxes and provide a work incentive for low- 
income taxpayers.[Footnote 5] It is a refundable federal income tax 
credit, meaning that qualifying working taxpayers may receive a refund 
greater than the amount of income tax they paid for the year. For tax 
year 2006, the maximum amount of EITC a taxpayer could receive was 
$4,536. 

Beginning in 1979, individuals could elect to receive the EITC in 
advance payments from their employer during the year along with their 
regular pay.[Footnote 6] One purpose of the advance payment is to 
provide employees with an immediate reward for their work effort rather 
than forcing them to postpone receiving the credit until they file 
their tax returns. To get the credit, at any time during the year, an 
employee would complete the Form W-5 and provide it to his or her 
employer.[Footnote 7] 

Upon receiving a completed Form W-5, the employer calculates the amount 
of the AEITC payment to include with the employee's pay by considering 
(1) the employee's wages, (2) whether the employee is married or 
single, and (3) if married, whether the employee's spouse has a Form W- 
5 in effect with an employer. The AEITC payment to the employee is 
considered to be equivalent to the employer making a payment to IRS for 
employee income tax withholding and employee and employer Social 
Security and Medicare tax. When employers file their quarterly tax 
returns, they show the total payments made to employees on the AEITC 
payment line on Form 941, "Employer's Quarterly Federal Tax Return." 
This amount is then subtracted from the total amount of tax the 
employer owes. 

At the end of the calendar year, the employer indicates the total AEITC 
payments the employee received on the employee's Form W-2. Employees 
are then required to report this amount either on their Form 1040 or 
Form 1040A tax return. Assuming the employee qualified for the EITC, 
the AEITC amount received should be reported on the tax return as other 
taxes, which, in effect, subtracts the amount received from the total 
amount of any EITC. If the employee did not qualify for the EITC, he or 
she is still required to file a tax return, regardless of income level, 
and the AEITC amount paid is added to any taxes owed. Figure 1 
illustrates this process and notes the major forms involved. 

Figure 1: AEITC Process for Employees and Employers: 

[See PDF for image] 

Source: GAO analysis of IRS data; IRS and PhotoDisc (images). 

[End of figure] 

An individual must meet certain requirements to qualify for the AEITC. 
Specifically, an individual must expect to (1) be able to claim the 
EITC for the current year (EITC requirements for 2006 are shown in 
table 1), (2) have at least one qualifying child, and (3) have earned 
income and adjusted gross income below a certain amount for that year. 
There are additional requirements that an individual may only have one 
Form W-5 in effect at a time and that he or she informs their employer 
if their spouse also has a Form W-5 in effect.[Footnote 8] AEITC 
recipients can receive up to 60 percent of the EITC benefits for one 
qualifying child. The maximum AEITC amount for 2006 was $1,648. 

Table 1: EITC Requirements for 2006 That All AEITC Recipients Must Also 
Expect to Meet: 

First, an individual must meet all the rules in this column; Second, an 
individual must meet all the rules in one of these columns, whichever 
applies; Third, an individual must meet the rules in this column. 

Rules for everyone: 

1. Adjusted gross income (AGI) must be less than: 

* $36,348 ($38,348 for married filing jointly) if more than one 
qualifying child.
 
* $32,001 ($34,001 for married filing jointly) if one qualifying child, 
or: 

* $12,120 ($14,120 for married filing jointly) if no qualifying child; 
First, an individual must meet all the rules in this column; 

Rules for everyone: 

2. Must have a valid Social Security number;

Rules for everyone: 

3. Filing status cannot be "Married filing separately."; 

Rules for everyone: 

4. Must be a U.S. citizen or resident alien all year; 

Rules for everyone: 

5. Cannot file Form 2555 or Form 2555-EZ (relating to foreign earned 
income); 

Rules for everyone: 

6. Investment income must be $2,800 or less; 

Rules for everyone: 

7. Must have earned income; 

Second, an individual must meet all the rules in one of these columns, 
whichever applies. 

Rules if have a qualifying child: 

8. Child must meet the relationship, age, and residency tests; 

Rules if have a qualifying child: 

9. Qualifying child cannot be used by more than one person to claim the 
EIC; 

Rules if have a qualifying child: 

10. Cannot be a qualifying child of another person; 

Rules if do not have a qualifying child: 

11. Must be at least age 25 but under age 65; 

Rules if do not have a qualifying child: 

12. Cannot be the dependent of another person; 

Rules if do not have a qualifying child: 

13. Cannot be a qualifying child of another person; 

Rules if do not have a qualifying child: 

14. Must have lived in the United States more than half of the year; 

Third, an individual must meet the rules in this column. 

Figuring and claiming the EITC: 

15. Earned income must be less than: 

* $36,348 ($38,348 for married filing jointly) if more than one 
qualifying child; 
* $32,001 ($34,001 for married filing jointly) if one qualifying child, 
or; 
* $12,120 ($14,120 for married filing jointly) if no qualifying child. 

Source: Excerpt of IRS Publication 596, Earned Income Credit. 

[End of table] 

A change in an individual's personal circumstances after submitting the 
Form W-5 could affect their eligibility for both the EITC and the 
AEITC. For example, an individual who received the AEITC, separated 
from his or her spouse during the year, and used the married filing 
separately filing status would not be eligible for the AEITC (or EITC). 
In such cases, when the individual files their tax return and reports 
the AEITC amount received, the amount would be added to any taxes due 
or subtracted from any refund. 

AEITC Use Low and May Be Difficult to Increase: 

About 3 percent of the EITC recipients potentially eligible for the 
advance, about 514,000 individuals on average, elected it in each year, 
tax years 2002 through 2004, according to data employers reported on 
the Form W-2. As shown in figure 2, about 21 million taxpayers received 
the EITC each year and of these recipients, about 17 million were 
eligible for the AEITC.[Footnote 9] 

Figure 2: Number of EITC Recipients, EITC Recipients Potentially 
Eligible for the AEITC, AEITC Recipients, and Total Dollars They 
Received, Tax Years 2002 through 2004: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

Note: This figure includes the valid, invalid name, and invalid number 
subpopulations. (see app. II). 

[End of figure] 

In total, AEITC recipients received an average of about $146 million in 
AEITC for each tax year from 2002 through 2004. Yet, those who elected 
it often received relatively few dollars from their employers. As table 
2 indicates, about half of all individuals who got AEITC received $100 
or less each year and about 75 percent received $500 or less. Even at 
$100 per year, this equates to about $8 a month, or $4 every 2 weeks, 
and even at $500 per year, this equates to about $42 a month or $19 
every 2 weeks. The amounts most individuals received were significantly 
less than the yearly maximum, and were consistent for the 3 years we 
reviewed. 

Table 2: Percentage of AEITC Recipients Receiving an Amount of AEITC, 
Tax Years 2002 through 2004: 

Amount of AEITC received: $1--$100; 
Percentage of AEITC recipients: 2002: 50; 
Percentage of AEITC recipients: 2003: 50; 
Percentage of AEITC recipients: 2004: 48. 

Amount of AEITC received: $101--$250; 
Percentage of AEITC recipients: 2002: 17; 
Percentage of AEITC recipients: 2003: 17; 
Percentage of AEITC recipients: 2004: 17. 

Amount of AEITC received: $251--$500; 
Percentage of AEITC recipients: 2002: 13; 
Percentage of AEITC recipients: 2003: 13; 
Percentage of AEITC recipients: 2004: 13. 

Amount of AEITC received: $501--$750; 
Percentage of AEITC recipients: 2002: 7; 
Percentage of AEITC recipients: 2003: 7; 
Percentage of AEITC recipients: 2004: 8. 

Amount of AEITC received: $751--$1,000; 
Percentage of AEITC recipients: 2002: 5; 
Percentage of AEITC recipients: 2003: 5; 
Percentage of AEITC recipients: 2004: 5. 

Amount of AEITC received: $1,001--yearly maximum; 
Percentage of AEITC recipients: 2002: 7; 
Percentage of AEITC recipients: 2003: 8; 
Percentage of AEITC recipients: 2004: 8. 

Source: GAO analysis of IRS data. 

Note: The AEITC yearly maximum was $1,503, $1,528, and $1,563 for tax 
years 2002, 2003, and 2004, respectively. This table includes the 
valid, invalid name, and invalid number subpopulations. 

[End of table] 

The number of individuals who have elected the AEITC has remained low 
for many years. For example, between 1990 and 1997, AEITC use never 
exceeded 2 percent of qualifying EITC recipients. IRS calculated use 
based on tax returns showing receipt of the AEITC divided by the EITC 
population that reported at least one qualifying child. Using this same 
methodology, use for tax years 2002 through 2004 was relatively the 
same at an average of 0.8 percent. Our figures for 2002 through 2004 
are higher than these prior AEITC figures because our figures are based 
upon the Forms W-2 that reported AEITC (see fig. 2). Historically, 
AEITC use has been based upon the number of individual federal tax 
returns that reported an amount on the AEITC line. The historic method 
excludes individuals who did not file a federal tax return and 
individuals who filed a federal tax return but did not report the 
AEITC. 

Additional demographic data about individuals who elected the AEITC are 
included in appendix III. These additional data represent new analysis 
that has not been previously available, including each recipient's 
filing method, age, and gender, and each employer's size. 

AEITC Use Has Not Increased Significantly Despite Several Targeted 
Efforts: 

Use of the AEITC has remained low for many years despite several 
targeted efforts to increase it. There have been several federal 
efforts targeted to increase AEITC use over the last approximately 15 
years, including both legislative and administrative changes. (A full 
description of these changes is included in app. IV.) One significant 
piece of legislation was OBRA '93, which involved the AEITC in two 
ways: it (1) reduced the maximum amount of EITC an individual could 
receive in advance and (2) required IRS to conduct outreach directly to 
potentially eligible AEITC recipients. First, OBRA '93 reduced the 
AEITC maximum from 100 percent to 60 percent of the maximum credit 
available to a taxpayer with one qualifying child. This change was made 
to improve compliance and lessen concerns that recipients would owe the 
difference when filing their federal tax return, which was thought to 
discourage AEITC use. 

Second, OBRA '93 directed IRS to send notices to taxpayers who were 
likely to be eligible for the AEITC for 2 years and directed the 
Secretary of the Treasury to study the effect of the notice program on 
AEITC use. Only some information is available about the first notice 
mailing, which occurred in 1994.[Footnote 10] For the first notice 
mailing, IRS mailed Publication 1235, "Advance Earned Income Tax Credit 
Brochure," and the Form W-5 to about 13.5 million taxpayers who were 
potential AEITC candidates during tax year 1993 informing them about 
the AEITC. AEITC use increased about 1 percent following this effort; 
however, because other outreach efforts were ongoing during this time, 
IRS could not conclude that the increase was attributable to the notice 
or any other effort specifically. 

In 1997, IRS mailed the notice to about 6 million taxpayers who claimed 
the EITC in tax year 1996, but did not report receiving AEITC on their 
federal tax return. With this second mailing, IRS created two groups, a 
test group of about 60,000 taxpayers that received the notice and a 
control group of about 60,000 taxpayers with similar characteristics 
who did not. Results from the IRS report indicated that about 1.27 
percent (771 taxpayers) of the tax returns in the test group reported 
the AEITC compared to 0.51 percent (309 taxpayers) of returns in the 
control group. The summary report concluded that further efforts to 
increase AEITC use substantially are unlikely to succeed.[Footnote 11] 
Further, the study recommended that notification of EITC recipients 
about the advance not be repeated. 

Some of IRS's administrative changes include outreach to specific 
groups and changes to publications. For example, after our 1992 report, 
the White House, the Treasury Department, and IRS conducted extensive 
EITC and AEITC outreach efforts, including a 1993 announcement of the 
AEITC by President Clinton. Other outreach efforts included IRS 
contacts with charitable, social welfare, and minority groups to 
encourage awareness of the EITC and AEITC among their memberships. IRS 
also contacted a number of employer organizations to encourage them to 
publicize the AEITC with their memberships. IRS also made changes to 
its forms, developed print and video products, and increased speaker 
seminars to inform the public about the AEITC. For example, IRS 
developed publicity materials, such as grocery bag and milk carton art, 
brochures, and posters; provided information in the Small Business 
Taxpayer Education Program guide; and increased outreach speaker 
seminar efforts.[Footnote 12] 

Presently, IRS continues to conduct outreach about the AEITC as part of 
its EITC outreach efforts. IRS focuses its outreach to large employer 
organizations or to specific large employers, which then promote the 
AEITC to employers or employees. 

Future AEITC Growth Is Unlikely: 

Increasing AEITC use in the future is unlikely for several reasons, but 
perhaps primarily because of potential recipients' preferences and high 
AEITC turnover. Interviews with IRS officials, other experts, and our 
prior AEITC work suggest that those eligible for the AEITC prefer 
receiving the EITC in a lump sum after they file their federal tax 
return instead of receiving relatively small portions spread throughout 
the year.[Footnote 13] Another reason is that, despite the reduction in 
the yearly AEITC maximum to 60 percent of the maximum credit available 
to a taxpayer with one qualifying child, results from an IRS-funded 
study using focus groups of EITC participants and interviews we 
conducted with experts indicated that potential recipients continue to 
have concerns that they would receive more AEITC than they were 
ultimately entitled to and that they would owe the difference when 
filing their federal tax return. 

In addition, AEITC growth is adversely affected by individuals who 
elect the AEITC and fail to elect it again, i.e., turnover. As table 3 
indicates, more than half of the individuals who elected the AEITC did 
so for the first time since 1999 in either tax year 2002, 2003, or 
2004. With the overall AEITC use remaining relatively constant over the 
3 years, this indicates the large number of yearly first-time 
recipients was almost equally offset by existing recipients forgoing 
the AEITC in a following year.[Footnote 14] The percentage of AEITC 
first-time recipients is much higher than the percentage of first-time 
EITC recipients, which is slightly less than one-third. Also, of the 
individuals who elected the AEITC for the first time in 2002 or 2003, 
about 28 percent elected it again in the following year, 2003 or 2004, 
respectively, while the remainder did not elect it again in the 
subsequent year. Conversely, only about 98,000 (9 percent) individuals 
elected the AEITC consecutively in all 3 years, 2002 through 2004. 

Table 3: Frequency of Individuals Electing the EITC and AEITC, Tax 
Years 2002 through 2004: 

EITC recipients; 
Tax year 2002: 20,918,823; 
Tax year 2003: 21,431,377; 
Tax year 2004: 21,721,218. 

First-time EITC recipients (percentage) [A, B]; 
Tax year 2002: 6,077,104: (29%); 
Tax year 2003: 5,313,089: (25%); 
Tax year 2004: 4,842,922: (22%). 

AEITC recipients; 
Tax year 2002: 531,799; 
Tax year 2003: 503,382; 
Tax year 2004: 507,957. 

First-time AEITC recipients (percentage) [A, C]; 
Tax year 2002: 271,600: (51%); 
Tax year 2003: 259,579: (52%); 
Tax year 2004: 259,282: (51%). 

First-time AEITC recipients who did not elect the advance in the 
subsequent year (percentage) [D]; 
Tax year 2002: 198,096: (73%); 
Tax year 2003: 187,871: (72%); 
Tax year 2004: N/A.

First-time AEITC recipients who elected the advance in the subsequent 
year (percentage) [D]; 
Tax year 2002: 73,504: (27%); 
Tax year 2003: 71,708: (28%); 
Tax year 2004: N/A. 

Number of individuals who consecutively received the AEITC (percentage) 
[E]; 
Tax year 2002: 97,998: (9%). 

Source: GAO analysis of IRS data. 

Note: N/A means not applicable. Tax year 2004 data are not applicable 
because we did not look at any subsequent years. All AEITC figures 
include the valid, invalid name, and invalid number subpopulations. 

[A] First-time EITC recipients are based upon how often a primary or 
secondary taxpayer filed a federal tax return receiving EITC for the 
first year since 1999. First-time AEITC recipients are based upon the 
first year an individual (defined by unique SSN) received AEITC since 
1999. IRS officials agreed this approach is a reasonable method to 
compare AEITC and EITC first-time use. 

[B] Percentage is out of EITC recipients. 

[C] Percentage is out of AEITC recipients. 

[D] Percentage is out of first-time AEITC recipients. 

[E] Number of individuals who consecutively received the AEITC are 
individuals who received the advance each year in tax years 2002 
through 2004. Percentage is of individuals who received the AEITC each 
year in tax years 2002 through 2004 out of all recipients who received 
it at least once during these same years (1,133,908 individuals). The 
percentage who received it twice during these years is 19 percent 
(213,234) and once is 73 percent (822,676). Percentages may not add due 
to rounding. 

[End of table] 

High Noncompliance Exists with AEITC Requirements: 

Overall, as many as 80 percent of all AEITC recipients did not comply 
with or made errors involving one of the three AEITC requirements that 
we reviewed, and they received about $282 million when the 3 years, 
2002 through 2004, are aggregated. Some taxpayers were noncompliant 
with more than one requirement. Those requirements are having a valid 
SSN, filing a federal tax return, and reporting the proper amount of 
AEITC received on the tax return (see fig. 3). 

Figure 3: Percentage of AEITC Recipients Compliant and Noncompliant 
with at Least One AEITC Requirement, Tax Years 2002 through 2004: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

Note: The pie chart includes the valid, invalid name, and invalid 
number subpopulations. 

[A] This includes the invalid name and invalid number subpopulations. 

[B] This includes the valid, invalid name, and invalid number 
subpopulations. 

[C] This is of individuals who filed a federal tax return in the valid, 
invalid name, and invalid number subpopulations. 

[End of figure] 

Specifically, in tax year 2002, we found that individuals were 
noncompliant with at least one of three requirements we reviewed 79 
percent of the time and they received about $93 million of AEITC. For 
tax year 2003 and 2004, individuals were noncompliant with at least one 
requirement 78 percent (about $91 million) and 79 percent (about $98 
million) of the time, respectively. 

Some of the noncompliance we identified could have resulted from IRS or 
employer clerical errors or improper reporting by the taxpayer. 
Therefore, some of the errors may be correctible or were corrected by 
filing an amended return. IRS cannot readily identify the number of 
amended returns specifically associated with the AEITC because such 
returns combine several credits onto one line. The explanation attached 
to the amended return would provide details on the change and any 
analysis of the explanation would be a manual process. 

About 20 Percent of AEITC Recipients Had an Invalid SSN: 

AEITC recipients are required to provide their employer with a valid 
SSN for the Form W-2.[Footnote 15] As table 4 illustrates, about 20 
percent (more than 100,000) of AEITC recipients each year may not have 
been eligible for the advance because they did not have a valid SSN on 
their Form W-2. Collectively, these individuals received between $37 
million and $39 million in AEITC each year. The data are consistent 
over the 3 years reviewed.[Footnote 16] 

Table 4: AEITC Recipients and Recipients with Invalid SSN on Form W-2, 
Individuals and Dollars, Tax Years 2002 through 2004: 

AEITC recipients; 
Tax year 2002: Individuals: 531,799; 
Tax year 2002: Dollars: $145 million; 
Tax year 2003: Individuals: 503,382; 
Tax year 2003: Dollars: $143 million; 
Tax Year 2004: Individuals: 507,957; 
Tax Year 2004: Dollars: $150 million.

AEITC recipients with an invalid SSN on Form W-2 (percentage)[A]; 
Tax year 2002: Individuals: 113,025: (21%); 
Tax year 2002: Dollars: $37 million: (26%); 
Tax year 2003: Individuals: 104,856: (21%); 
Tax year 2003: Dollars: $37 million: (26%); 
Tax year 2004: Individuals: 109,622: (22%); 
Tax year 2005: Dollars: $39 million: (26%). 

Source: GAO analysis of IRS data. 

[A] Percentage is out of AEITC recipients. This analysis includes the 
invalid name and invalid number subpopulations. 

[End of table] 

For purposes of this report, invalid SSNs include instances where the 
SSN did not match SSA's records (i.e., the number was never assigned by 
SSA) and the SSN/name combinations reported on the Form W-2 did not 
match SSA records.[Footnote 17] Some of these individuals were likely 
eligible for the AEITC. For example, a name/SSN mismatch could include 
instances when a woman who receives the AEITC marries and changes her 
name with her employer but not with SSA. This could result in the 
employer issuing a Form W-2 in the new name, but IRS and SSA only 
identifying her by the former name.[Footnote 18] 

Individuals who file a federal tax return are required to include a 
valid SSN on their return.[Footnote 19] An individual who provides an 
invalid SSN on the tax return is not compliant in meeting AEITC 
requirements and may also violate the Social Security Act.[Footnote 20] 
Such an individual is also required to provide a valid SSN to their 
employer for income tax withholding purposes and for purposes of 
certifying eligibility for the AEITC and could be subject to a penalty 
for failure to do so.[Footnote 21] Further, if the individual does not 
file a valid SSN, IRS is unable to assess the recipient's federal tax 
liability and SSA cannot credit the recipient for money withheld for 
Social Security purposes.[Footnote 22] Also, because taxpayers often do 
not report receipt of the advance on their tax return, IRS cannot 
determine whether the taxpayer owes money or deserves a refund. 

Almost 40 Percent of AEITC Recipients Did Not File the Required Federal 
Tax Return: 

All AEITC recipients are required to file a federal tax return, 
regardless of the amount of their income, which is generally the 
primary basis for determining whether a return is required to be filed. 
Table 5 shows that between 36 and 40 percent, about 200,000 AEITC 
recipients, did not file a required federal tax return each year. 
Collectively, these individuals received between $42 million and $50 
million of AEITC benefits.[Footnote 23] 

Table 5: AEITC Recipients and Recipients Who Did Not File a Tax Return, 
Individuals and Dollars, Tax Years 2002 through 2004: 

AEITC recipients; 
Tax year 2002: Individuals: 531,799; 
Tax year 2002: Dollars: $145 million; 
Tax year 2003: Individuals: 503,382; 
Tax year 2003: Dollars: $143 million; 
Tax year 2004: Individuals: 507,957; 
Tax year 2004: Dollars: $150 million. 

AEITC recipients who did not file a federal tax return (percentage)[A]; 
Tax year 2002: Individuals: 192,857: (36%); 
Tax year 2002: Dollars: $42 million: (29%); 
Tax year 2003: Individuals: 186,423: (37%); 
Tax year 2003: Dollars: $44 million: (31%); 
Tax year 2004: Individuals: 200,706: (40%); 
Tax year 2004: Dollars: $50 million: (33%). 

Source: GAO analysis of IRS data. 

[A] Percentage is of AEITC recipients. This analysis includes the 
valid, invalid name, and invalid number subpopulations. 

[End of table] 

About 56,000 to 60,000 of the about 200,000 individuals who did not 
file the required tax return (about 30 percent) had an invalid SSN on 
the Form W-2 each year, as shown in table 6. Having a valid SSN is 
another AEITC requirement, discussed previously, which means these 
individuals were noncompliant or made an error with at least two AEITC 
requirements. 

Table 6: AEITC Recipients, Recipients Who Did Not File a Federal Tax 
Return, and Recipients Who Did Not File a Tax Return and Had an Invalid 
SSN on the Form W-2, Individuals and Dollars, Tax Years 2002 through 
2004: 

(Continued From Previous Page) 

AEITC recipients; 
Tax year 2002: Individuals: 531,799; 
Tax year 2002: Dollars: $145 million; 
Tax year 2003: Individuals: 503,382; 
Tax year 2003: Dollars: $143 million; 
Tax year 2004: Individuals: 507,957; 
Tax year 2004: Dollars: $150 million. 

AEITC recipients who did not file a federal tax return (percentage) 
[A]; 
Tax year 2002: Individuals: 192,857: (36%); 
Tax year 2002: Dollars: $42 million: (29%); 
Tax year 2003: Individuals: 186,423: (37%); 
Tax year 2003: Dollars: $44 million: (31%); 
Tax year 2004: Individuals: 200,706: (40%); 
Tax year 2004: Dollars: $50 million: (33%). 

AEITC recipients who did not file a federal tax return and had an 
invalid SSN on Form W-2 (percentage) [B]; 
Tax year 2002: Individuals: 59,607: (31%); 
Tax year 2002: Dollars: $20 million: (48%); 
Tax year 2003: Individuals: 55,601: (30%); 
Tax year 2003: Dollars: $20 million: (45%); 
Tax year 2004: Individuals: 58,370: (29%); 
Tax year 2004: Dollars: $21 million: (42%). 

Source: GAO analysis of IRS data. 

[A] Percentage is of AEITC recipients. This analysis includes the 
valid, invalid name, and invalid number subpopulations. 

[B] Percentage is of AEITC recipients who did not file a federal tax 
return. An unknown portion of these recipients may have corrected the 
SSN error and subsequently filed a tax return. This analysis includes 
the invalid name and invalid number subpopulations. 

[End of table] 

There are several reasons why a significant number of AEITC recipients 
might not have filed a federal tax return. For example, depending on 
their filing status, age, and type of income they receive, recipients 
may not have had a filing responsibility other than for the AEITC and 
they may not have remembered or understood they must file a return. In 
addition, AEITC recipients may not have filed because they were not 
initially eligible or they became ineligible for the AEITC because of a 
change in their personal circumstances, and filing would require them 
to pay back the AEITC they received. 

When individuals are required to file a federal tax return and do not, 
IRS cannot readily identify whether the individual was eligible for the 
advance or whether they owed IRS any of the amounts they received. 
Conversely, by not filing a federal tax return, some individuals did 
not receive additional EITC monies that they could only receive had 
they filed. 

About Two-Thirds of AEITC Recipients Who Filed a Federal Tax Return 
Misreported the Amount of AEITC They Received: 

All AEITC recipients are required to report on their federal tax return 
the amount of AEITC they received according to the Form(s) W-2. 
Reporting this amount allows the IRS to determine whether the taxpayer 
received too much AEITC, and owes money back to the IRS, or whether the 
taxpayer is entitled to additional amounts of the EITC. 

Of the approximately 60 percent (about 300,000) AEITC recipients who 
filed a federal tax return, two-thirds misreported the amount they 
received in tax year 2002 through 2004, as shown in table 7. 
Misreported means that the total amount of AEITC reported on the Form W-
2 does not match the AEITC amount reported on the federal tax return. 
Approximately one-third of the federal tax returns correctly matched to 
the Form(s) W-2 AEITC amount. Of those that misreported, the vast 
majority did not report receiving any AEITC. 

Table 7: AEITC Matches and Mismatches between Form(s) W-2 and Filed Tax 
Returns, Tax Years 2002 through 2004: 

(Continued From Previous Page) 

Number of individuals who filed federal tax returns[A]; 
Tax year 2002: Number: 342,841; 
Tax year 2002: Dollars: $108,189,196; 
Tax year 2003: Number: 320,588; 
Tax year 2003: Dollars: $101,078,453; 
Tax year 2004: Number: 309,724; 
Tax year 2004: Dollars: $102,512,732. 

Number of returns that matched the Form(s) W-2 AEITC amount 
(percentage)[B]; 
Tax year 2002: Number: 112,421: (33%); 
Tax year 2002: Dollars: $51,457,254; (48%); 
Tax year 2003: Number: 111,266: (35%); 
Tax year 2003: Dollars: $52,246,111: (52%); 
Tax year 2004: Number: 106,980: (35%); 
Tax year 2004: Dollars: $52,338,650: (51%). 

Number of mismatches between Form(s) W-2 and filed federal tax return 
(percentage)[B]; 
Tax year 2002: Number: 230,420: (67%); 
Tax year 2002: Dollars: $56,731,942: (52%): 
Tax year 2003: Number: 209,322: (65%); 
Tax year 2003: Dollars: $48,832,342: (48%); 
Tax year 2004: Number: 202,744: (65%); 
Tax year 2004: Dollars: $50,174,082: (49%). 

Number of mismatches due to nonreporting of AEITC (percentage)[C]; 
Tax year 2002: Number: 222,691: (97%); 
Tax year 2002: Dollars: $53,343,102: (94%); 
Tax year 2003: Number: 202,152: (97%); 
Tax year 2003: Dollars: $45,487,033: (93%); 
Tax year 2004: Number: 195,571: (96%); 
Tax year 2004: Dollars: $46,806,223: (93%). 

Number of mismatches due to underreporting of AEITC (percentage)[C]; 
Tax year 2002: Number: 4,996: (2%); 
Tax year 2002: Dollars: $2,349,197: (4%); 
Tax year 2003: Number: 4,779: (2%); 
Tax year 2003: Dollars: $2,389,854: (5%); 
Tax year 2004: Number: 4,881: (2%); 
Tax year 2004: Dollars: $2,337,603: (5%). 

Number of mismatches due to overreporting AEITC (percentage)[C]; 
Tax year 2002: Number: 2,733: (1%); 
Tax year 2002: Dollars: $1,039,643: (2%); 
Tax year 2003:  Number: 2,391: (1%); 
Tax year 2003: Dollars: $955,455: (2%); 
Tax year 2004: Number: 2,292: (1%); 
Tax year 2004: Dollars: $1,030,256: (2%). 

Source: GAO analysis of IRS data. 

Note: Percentages may not add due to rounding. This table includes 
individuals who filed a tax return in the valid and invalid name 
subpopulations. 

[A] The total number of filed federal tax returns is greater than the 
total number of AEITC recipients who filed a return because there were 
some dependents who received AEITC and were listed on more than one 
federal tax return. The total number of AEITC recipients who filed a 
federal tax return can be calculated using table 5 by subtracting the 
number of AEITC recipients who did not file a federal tax return from 
the number of AEITC recipients. 

[B] We considered a match to be anything plus or minus a dollar on the 
Form W-2 in order to allow for taxpayer rounding. We tested the 
sensitivity of this result by using a difference between the AEITC 
reported on the Form W-2 and the tax return of up to $100 and the 
results were similar. Percentage is of number of individuals who filed 
federal tax returns. 

[C] Percentage is of number of mismatches between Form(s) W-2 and filed 
federal tax returns. 

[End of table] 

Taxpayers may not report the amount of AEITC they received because they 
either forget or do not know they are required to do so. Underreporting 
can occur when there is a computation error involving multiple Forms W- 
2, a taxpayer disagrees with the amount reported on the Forms W-2, or 
there is willful noncompliance. IRS officials in the AUR program pursue 
some AEITC cases where taxpayers either underreport or do not report 
receipt of the AEITC. These procedures are discussed in detail in a 
following section. 

Due to the high number of mismatches, many of the AEITC recipients who 
also claimed the EITC likely received excess benefits. For example, of 
the 222,691 taxpayers who did not report receipt of the AEITC on their 
tax return in tax year 2002, almost half went on to claim the EITC. We 
determined that those taxpayers received nearly $22 million in excess 
AEITC benefits. For the 3 years, 2002 through 2004, taxpayers received 
a total of about $64 million in excess AEITC benefits. 

AEITC Recipients Were Noncompliant with Other Program Requirements: 

Additional noncompliance or errors existed with other program 
requirements, such as receiving excess AEITC. Details and demographic 
information on the noncompliant individuals are in appendix V. 

IRS's AEITC Compliance Procedures Have Limited Effectiveness and 
Options to Better Address Noncompliance Have Advantages and 
Disadvantages: 

IRS's Submission Processing is responsible for receiving, processing, 
and archiving the nation's federal tax returns, payments, and 
information returns. In the context of AEITC, Submission Processing 
attempts to catch mismatches between both paper and electronic tax 
returns and Forms W-2 (see fig. 4). 

Figure 4: Submission Processing Procedures for Paper and Electronic Tax 
Returns Reporting AEITC: 

[See PDF for image] 

Source: GAO analysis of IRS procedures. 

[A] After some initials checks, such as comparing the EITC/AEITC line 
of a tax return to the Form W-2(s), all paper returns go to Submission 
Processing staff, who enter select data from the Form 1040 into an 
electronic database. 

[B] Effective January 2, 2007, at our suggestion, Submission Processing 
changed its selection criteria to include filing status. 

[End of figure] 

Between 32 percent and 22 percent of tax returns reporting an AEITC 
amount were filed on paper between tax years 2002 and 2004, 
respectively. If a paper tax return has an entry on either the AEITC or 
EITC lines on the Form 1040, Submission Processing tax examiners are 
required to ensure that the AEITC amount reported on the return matches 
the amount in box 9 of the Form(s) W-2, which records the amount of 
AEITC paid by the employer to the employee.[Footnote 24] 

If the amounts on the AEITC line of the tax return and Form W-2(s) 
match, the tax examiner takes no further action. If the amounts differ 
(e.g., the tax return reports a lesser amount than is reflected on the 
Form(s) W-2) the examiner is required to adjust the entry on the return 
to equal the total AEITC amount from the Form(s) W-2. All paper returns 
then go to staff who enter data from the Form 1040 into an electronic 
database. 

After an examiner makes an adjustment, IRS sends a letter to the 
taxpayer explaining that an adjustment was made and it was based on the 
mismatch between the tax return and the Form(s) W-2. IRS sent 282 and 
220 such letters in tax years 2003 and 2004, respectively.[Footnote 25] 
Submission Processing's role is to ensure that the return amount is 
consistent with the Form W-2, not to determine which of the differing 
numbers accurately reflects the amount of AEITC actually paid to the 
employee. If the taxpayer disagrees with the adjustment, e.g., believes 
that the amount on the Form(s) W-2 is incorrect, the taxpayer can 
dispute it. 

Between 68 percent and 78 percent of returns reporting an AEITC amount 
were filed electronically between tax years 2002 and 2004. Submission 
Processing runs a computer check to find any mismatches between the 
amounts on the electronic tax returns and the electronic Form(s) W-2. 
When mismatches are found, Submission Processing rejects the return and 
sends it to the taxpayer or preparer to correct and retransmit to IRS. 

In most tax preparation software, once a user enters an amount from the 
Form(s) W-2, the amount is automatically transferred to the appropriate 
line of the tax return. Thus, if the user errs in entering the proper 
amount from the Form W-2, the software would enter this erroneous 
number on the tax return. As a result, Submission Processing rarely 
identifies AEITC tax return/Form(s) W-2 mismatches because the original 
Form(s) W-2 from the employer(s) is rarely included in electronic 
filings.[Footnote 26] IRS rejected 172 electronically filed returns 
reporting AEITC in tax year 2004 and 147 in tax year 2003. Adding these 
mismatches to the paper return mismatches, Submission Processing found 
392 mismatches in tax year 2004 and 429 in tax year 2003. 

Next, Submission Processing sends the return through its Error 
Resolution System (ERS) when the AEITC amount on the Form 1040 meets 
certain selection criteria.[Footnote 27] If ERS finds that AEITC 
exceeds this amount, a tax examiner matches the Form 1040 AEITC amount 
to the Form W-2(s). If they match, no action is taken, and the return 
is posted in IRS's Masterfile, the agency's central repository for 
taxpayer information. If there is a mismatch, IRS adjusts the return to 
match the Form(s) W-2, and IRS sends a letter to the taxpayer 
describing the error and the ERS correction. Taxpayers who disagree 
with the change can dispute it. The ERS process thus serves as a back- 
up check in case the earlier physical or electronic processes missed 
these AEITC mismatches. ERS examined 3,380 tax returns with AEITC in 
tax year 2004. IRS was not able to tell us how many of these returns 
involved mismatches. 

Prior ERS selection criteria would not have identified some returns for 
individuals when the Form(s) W-2 had AEITC amounts that were above the 
legal maximums, but below the ERS selection criteria. While 
interviewing Submission Processing officials, we suggested that the 
criteria be modified and associated with filing status. IRS made such 
changes to the criteria, effective January 2, 2007. 

IRS's Procedures to Verify SSNs Have Limited Effectiveness for AEITC 
Recipients Due to Taxpayer Noncompliance in Reporting and Filing: 

Submission Processing's procedures have limited effectiveness in 
verifying that AEITC recipients have a valid SSN because, as previously 
noted, many individuals do not file the required tax return and, for 
those who do, most do not report receipt of the advance. 

For both paper and electronically filed returns, Submission Processing 
checks IRS's Data Master (DM-1) file, which is a database that 
includes, among other things, all validly issued SSNs and the 
individual's name associated with each SSN. Submission Processing 
rejects electronically filed tax returns with an invalid SSN on the 
Form 1040, including those from taxpayers who received the AEITC, and 
sends back the tax return to the taxpayer for correction before 
processing. For returns filed on paper with an invalid SSN on the Form 
1040, Submission Processing processes the return, but disallows certain 
credits and exemptions, such as the EITC. If the taxpayer's SSN is 
invalid and the AEITC is claimed on a paper return, Submission 
Processing processes any AEITC reported. Since the AEITC is reported as 
tax, it will either offset all or part of any refund due or, if no 
refund is due, require the taxpayer to pay back the full AEITC amount. 

Thus, of the approximately 514,000 individuals who received the AEITC 
each year between tax years 2002 through 2004, Submission Processing's 
procedures would apply to about 118,000 taxpayers--those who filed and 
reported receipt of the advance. For nonfilers, Submission Processing 
cannot verify SSNs because there are no tax returns---the basis of its 
examination. While Submission Processing performs a SSN verification 
for those who file and do not report receipt of the AEITC, it is not 
effective for the advance since the advance is not reported on the 
return. 

Many Taxpayers Underreported AEITC and IRS Worked on a Limited Number 
of These Cases, Some of Which Submission Processing Did Not Find: 

After a return has been processed, the next point when IRS might 
identify and correct AEITC noncompliance is when it matches tax returns 
to other documents it receives. Overall, each year AUR identifies about 
14 million discrepancies between taxpayer income and deduction 
information submitted by third parties and amounts reported on 
individual income tax returns. AUR has the resources to only work on a 
fraction of these cases each year and uses criteria, such as revenue 
collection potential, for case selection. 

For AEITC, AUR compares the amount of AEITC that employers report on a 
taxpayer's Form W-2 to the amount reported on an individual's tax 
return (see fig. 5). AUR receives this information in separate 
databases from SSA and undertakes these comparisons in August and 
December of each year--well after Submission Processing has completed 
its review of the returns and associated Forms W-2. 

Figure 5: AUR Program Procedures to Resolve Discrepancies between AEITC 
Amounts on Forms W-2 and Tax Returns: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

[End of figure] 

Generally, AUR does not take action when the AEITC amounts match the 
amounts reported on the return or when the discrepancy does not meet 
IRS case selection criteria. For example, for tax year 2003, our data 
show that there were about 209,000 tax returns where the AEITC amount 
did not match the amount on taxpayers' Forms W-2, primarily because the 
returns failed to report any AEITC or underreported the amount. AUR 
identified about 25,000 of those cases in tax year 2003 as potential 
cases on which to work. 

In tax year 2003, Submission Processing identified 429 instances where 
the amount of AEITC reported on the tax return was less than the amount 
on the Forms W-2, which was much fewer than the about 25,000 mismatches 
detected by AUR that year. The mismatches involving individuals who 
filed paper returns and reported receiving the AEITC should have been 
caught by Submission Processing because its examiners are supposed to 
match the amount reported on the return with the Form(s) W-2. However, 
Submission Processing examiners review huge volumes of returns and it 
can be a challenge to identify the relatively few with AEITC. 

Because AUR does not break down its AEITC cases by filers who 
underreport and filers who fail to report any AEITC amount or by filers 
of paper versus electronic returns, it is not possible to determine 
exactly how many AUR cases should have been caught by Submission 
Processing. AUR officials told us that it would be difficult for them 
to routinely break out this information in this way because this would 
require new computer programming and the budget for new programming 
requests was reduced for fiscal year 2007. 

In accordance with return processing procedures, if Submission 
Processing identifies an underreporter, it reduces the size of the 
refund that will be sent to the taxpayer. Thus, using Submission 
Processing, rather than AUR, better protects revenue because the 
erroneous EITC amounts are never paid to the taxpayer. If AUR works on 
the case, it assesses the tax owed, but further IRS action is required 
to collect the actual overpayment--through a future refund offset or a 
current effort to collect the refund paid erroneously. 

Although catching AEITC underreporting in Submission Processing would 
better protect revenue than do AUR processes, the benefits to improving 
Submission Processing to catch AEITC errors may already be small and 
could get smaller. This is because the number of EITC returns filed 
electronically is more than 68 percent and has been growing and, as 
previously noted, Submission Processing identifies few mismatches for 
returns that are filed electronically because the electronic Form W-2 
is an iteration of the user's entries. In its July 2006 report, the 
Treasury Inspector General for Tax Administration recommended that IRS 
reemphasize the use of the current AEITC review procedures for paper 
returns, but agreed with IRS that it could not implement additional 
procedures for electronic returns because IRS is unable to fully verify 
the accuracy of the Form(s) W-2 during electronic processing of 
returns. 

AUR worked on about twice as many AEITC cases in tax year 2003 as in 
the previous year. The total amount of tax assessed in these cases in 
tax year 2003 was more than three times the amount in the previous 
year, and the amount assessed per case increased about 71 percent, from 
$555 to $947, which is near the average AUR assessment of about 
$1,000.[Footnote 28] These trends were either holding steady or 
improving for the 84 percent of tax year 2004 cases for which we had 
data when we completed our review. The number of cases in which the 
taxpayer fully agreed with the assessment rose from about 29 percent in 
tax year 2002 to about 41 percent in tax year 2003. The number of cases 
in which IRS withdrew the assessment, generally after the taxpayer 
provided documentation that it was erroneous, dropped from about 15 
percent in tax year 2002 to about 4 percent in tax year 2003. 

IRS Works on a Limited Number of AEITC Nonfiler Cases: 

As previously noted, from tax years 2002 through 2004, about 40 percent 
of the individuals who had a Form W-2 reporting that they received 
AEITC did not file a tax return as required by law. 

All nonfiler cases, including AEITC nonfiler and other lower dollar 
nonfiler cases, are eligible to be worked on by IRS's Wage and 
Investment and Small Business/Self Employed divisions. IRS's policy 
dictates that some lower dollar cases and even some cases in which 
nonfilers are due a refund may be worked on to ensure that all kinds of 
cases have the possibility of being worked on. Still, a key criterion 
that IRS uses to determine cases on which to work is potential revenue 
or the anticipated net balance due. The higher this amount, the more 
likely the case will be worked on. IRS does not track the total number 
of nonfiler cases worked on or the kinds of income or credits taken by 
the nonfilers whose cases were worked on. Because IRS selects cases 
based on these criteria, IRS officials said they worked on few AEITC 
nonfiler cases. 

Beyond AEITC nonfiler cases potentially being worked on by Wage and 
Investment and the Small Business/Self Employed divisions, these cases 
are also eligible to be worked on by other IRS programs. For example, 
the automated collection system involves calls from IRS staff to 
taxpayers asking them to file a return or explain why they believe 
filing is not required.[Footnote 29] IRS's automated substitute for 
return program involves collection staff preparing a tax return on 
behalf of a nonfiler based on third-party and other information that 
IRS has available. Once a return has been created, IRS can act to 
collect any taxes due. 

During 2005 through 2006, IRS conducted a test to determine whether 
receipt of AEITC should be a criterion to determine nonfiler cases on 
which to work. The test involved working on cases from tax years 2000 
through 2003 for 433 taxpayers drawn from a sample of taxpayers who did 
not file a return in tax year 2002, but did receive the AEITC that 
year, and whose income was between $35,000 and $50,000 for that 
year.[Footnote 30] The cases were worked on in the same way that IRS 
works on other nonfiler cases. 

In interviews with IRS officials about the AEITC nonfiler test, we 
found that the agency's test plan lacked documentation and detail, such 
as test justification, likely costs and benefits, and implementation 
details. It also lacked a rationale for some important decisions 
underlying the test and some changes implemented after the test began 
were also not well documented. In our report on three tests conducted 
by IRS in 2004 to address leading sources of EITC errors, we noted that 
the lack of such documentation hindered monitoring and oversight and 
did not foster a common understanding of the tests among management and 
staff.[Footnote 31] One of our recommendations was that the rationale 
for key decisions on such tests be documented, and the Commissioner of 
Internal Revenue agreed with that recommendation. 

Because IRS had not completed its evaluation of the test as of June 
2007, the agency has not decided whether AEITC should be a criterion to 
determine nonfiler cases on which to work. Regardless of the agency's 
decision, however, it is unlikely to significantly reduce the number of 
AEITC nonfilers because any increase in cases worked on would likely 
represent only a small number of such nonfilers, relative to both the 
total number of AEITC nonfilers and all nonfiler cases worked on by 
IRS. 

Collection Handles Few AEITC Cases: 

All AEITC cases are eligible to be worked on in IRS's collection 
program; however, it generally does not work on cases involving AEITC 
because the amounts involved are below selection criteria that 
determine which cases to pursue. 

Collection does not keep track of the specific types of income and 
credits claimed by the individuals whose cases they handle. Still, 
collection officials stated that they worked on few, if any, cases 
involving the AEITC. Instead, these cases go into "deferral status," 
which means any refund the taxpayer is due will be reduced until the 
balance due has been paid off. Like other taxpayers with an outstanding 
tax liability, these taxpayers would also get a notice stating how much 
they owe. 

Collection had an effort under way to work on cases under its selection 
criteria by setting up automatic monthly installment arrangements and 
notifying taxpayers that they were expected to begin paying what they 
owed. Officials said that too many taxpayers ignored the arrangements, 
and as a result, the program was determined to be cost prohibitive. The 
program was therefore discontinued. 

IRS Audits and Criminal Investigation Identified a Small Number of 
AEITC Cases: 

IRS annually audits about 500,000 of the more than 21 million tax 
returns that claim the EITC. With only about 3 percent of EITC 
recipients potentially eligible for the advance receiving it, only a 
small number of the audited returns are likely to involve the 
AEITC.[Footnote 32] When IRS audits tax returns claiming the EITC, an 
examiner is required to determine if the taxpayer was eligible for it, 
whether the taxpayer took the AEITC and, if so, whether the taxpayer 
reported the correct amount. If the examiner determines that the 
taxpayer was not eligible for the EITC, then the AEITC is disallowed. 

In addition, Criminal Investigation, which investigates potential 
criminal violations of the tax code and related financial crimes, has 
identified six cases associated with the AEITC since 2001. Five of the 
six cases involved refund fraud based on individuals creating fake 
businesses in order to obtain the AEITC. The other case involved a 
business owner attempting to evade employment tax by falsely signing up 
employees for the AEITC, but not including the credit in their 
paychecks. 

Soft Notices Have Increased Compliance in Situations Similar to AEITC, 
but Turnover and Difficulty Locating Recipients May Limit Effectiveness 
for AEITC: 

Because IRS's enforcement resources cannot fully cover all areas of 
noncompliance, including AEITC noncompliance, the agency has tried to 
cost effectively increase voluntary compliance in some areas that 
involve relatively small amounts of money by mailing taxpayers soft 
notices. Soft notices are letters that ask taxpayers to comply with a 
certain requirement in the future or, if the notice informs them that 
they are not entitled to a benefit that they received, to file an 
amended return. 

While IRS officials said there have not been any soft notices 
specifically targeting AEITC noncompliance, we reviewed the results of 
three tests that involved taxpayers who were not filing accurately-- 
many of whom would not otherwise be subject to an enforcement action. 
Additionally, IRS has modified one soft notice test that includes cost 
estimates. Each of the completed soft notice efforts show some benefits 
in improving compliance, however, they may be less effective for AEITC 
recipients. 

The First Soft Notice Test: The first soft notice--called the 
"Duplicate TINs" test--involved different taxpayers claiming the same, 
or a duplicate, TIN for a dependent or qualifying child in order to 
obtain an exemption, the EITC, or child tax credit benefits. In tax 
year 2002, IRS identified a total of about 2.4 million taxpayers who 
used a duplicate TIN. IRS sent soft notices to about 820,000 taxpayers. 

In November 2005, IRS reported that after receiving the soft notice, 
11.4 percent of the population amended their tax year 2002 
returns.[Footnote 33] Other results focused on taxpayers who received 
the notice for tax year 2002 and whether they repeated the use of a 
duplicate TIN on their tax years 2003 and 2004 tax returns. The results 
were as follows: 

* 84.9 percent did not repeat their behavior in either of the ensuing 
years; 

* 7.7 percent repeated the behavior in 2003, but not again in 2004; 

* 4.0 percent did not repeat the behavior in 2003, but did so in 2004; 
and: 

* 3.4 percent repeated the behavior for both ensuing years. 

Although IRS did not report the costs associated with this test, it did 
estimate the revenue that would have been lost without the soft 
notices. IRS reported that it protected a total of $218.3 million using 
the Duplicate TINs test. Due to limitations in the research design, 
such as not using a control group, IRS reported that it was uncertain 
whether these results were solely influenced by the receipt of a soft 
notice or if other factors may have contributed to the change in 
taxpayer behavior and subsequent revenue protected. IRS no longer 
considers this a test and continues to send out soft notices for 
Duplicate TINs issued each year. 

The Second Soft Notice Test: The second soft notice test--called the 
"AUR Soft Notice" test--involved filers who underreported small amounts 
of certain categories of income, such as wages, unemployment insurance, 
or sales of securities. In December and January 2004 and 2005, 
respectively, IRS sent 500 soft notices to randomly selected taxpayers 
who underreported income on their tax year 2003 returns. IRS also 
randomly selected a control group of 500 taxpayers who underreported 
small amounts, but did not send notices to this group. 

An outside consultant that IRS hired to determine the effectiveness of 
the test reported in October 2005 that (1) soft notices appeared to 
have a beneficial result in reducing repeat behavior and (2) IRS 
resources were not overburdened by the notices.[Footnote 34] Their 
conclusion was based on several results. First, after receiving the 
soft notice, 71 out of the 500 taxpayers (14.2 percent) filed an 
amended return. Second, only 33 taxpayers (6.6 percent) who received 
the notice repeated the underreporting the following year. In contrast, 
174 taxpayers in the control group (34.8 percent) repeated their 
underreporting. Third, the consultants did not consider IRS resources 
to be burdened because only 45 of 500 taxpayers (9 percent) called IRS 
with questions. Similarly, the study found there was limited 
undeliverable mail--only for 3 taxpayers (0.6 percent). An additional 
test was conducted for fiscal year 2006 and had similar positive 
results. IRS is in the process of determining whether it will send out 
soft notices for AUR in the future. 

The Third Soft Notice Test: The third soft notice test--called the 
"Dependent Database" test--involved cases selected for three EITC 
related issues, including qualifying child, filing status, and Schedule 
C, "Profit or Loss from Business," errors. IRS found that 2.4 million 
taxpayers appeared to have had errors on their tax returns. In November 
2005, IRS selected about 12,500 taxpayers to determine the impact of 
soft notices on taxpayers' behavior when filing their tax year 2005 
return. About another 12,500 taxpayers were selected as a control group 
not to receive the notice. 

In its October 2006 report, IRS found that, although there was a 
difference between the test group of taxpayers who received the soft 
notice and the control group that did not, the direct relationship 
between receiving a soft notice and taxpayers' subsequent filing 
behavior was weak.[Footnote 35] Specifically, the report cited that 88 
percent of the test group and 86 percent of the control group changed 
their subsequent tax year filing behavior, including not breaking the 
same rule, amending the prior year return, or not filing a 2005 return. 
Specific noteworthy results were: 

* 84 percent of the test group and 83 percent of the control group 
filed a return in the subsequent year; 

* 46 percent of the test group and 44 percent of the control group 
broke no rules at all; 

* 26 percent of the test group and 25 percent of the control group 
broke a different rule; 

* 12 percent of the test group and 14 percent of the control group 
repeated their behavior the next year by breaking the same rule; and: 

* 1 percent of the test group and 0.4 percent of the control group 
amended their prior-year return. 

Also in October 2006, IRS modified the Dependent Database test in both 
the Wage and Investment and Small Business/Self Employed divisions to 
target notices to another population, i.e., noncustodian person(s) 
claiming a child. IRS prepared a preliminary cost analysis for this 
soft notice test based on a sample of 300,000 taxpayers. It estimated 
the total costs of sending out 300,000 soft notices to be about 
$533,000, which included $449,000 for the labor to process amended 
returns and answer telephone calls and $84,000 for mailing. Additional 
information, including the results of this test, was not available as 
of mid-June 2007. 

Although IRS did not develop criteria for these soft notice tests about 
what would constitute a success, such as a self-correction percentage, 
an IRS official knowledgeable about the tests said the agency considers 
the three completed tests a success, despite the few shortcomings. The 
first and second tests were considered successes because they led to 
noteworthy changes in taxpayer behavior. The third test was considered 
a success because, although taxpayer behavior did not change 
significantly, officials considered it a cost-effective way to have an 
enforcement presence among these taxpayers. Officials thought that a 
soft notice test could be beneficial for reducing AEITC noncompliance 
as well, particularly since the amount of money involved with AEITC is 
low and the noncompliance might not otherwise be addressed by IRS. 

Although soft notices may have some potential to address certain AEITC 
noncompliance, characteristics of the AEITC population might make such 
notices less effective or more costly than for the test populations for 
two reasons. First, AEITC turnover is high. In each tax year 2002 
through 2004, more than half of the individuals were first time 
recipients. Moreover, about 73 percent of first-time AEITC recipients 
in tax years 2002 and 2003 did not elect the AEITC the following year 
and, thus, would not repeat noncompliance related to the 
AEITC.[Footnote 36] Second, almost 40 percent of AEITC recipients do 
not file a tax return, which means that IRS may not have a current 
address for those taxpayers. If IRS were to send soft notices to AEITC 
nonfilers using the last known address, a significant number of 
individuals may no longer reside there. This means IRS might not be 
able to locate them or it might spend additional resources trying. 

Federal On-Line SSN Verification Services Could Be Used to Determine 
AEITC Eligibility, but IRS and SSA Have Raised Concerns: 

More than 100,000 AEITC recipients had invalid SSNs and reported 
receiving millions of dollars in total benefits for each tax year 2002 
through 2004 without any substantial check of their eligibility. 
Because of the low-dollar amounts involved per taxpayer, IRS worked on 
only a small number of these cases. IRS does not have an up-front 
control or procedure in place to require employers to verify that an 
employee seeking the AEITC has a valid SSN, which could address this 
noncompliance. Two federal on-line services have the potential to be 
used to implement such controls. Although the services could be used 
for this purpose, IRS and SSA officials raised several concerns about 
implementing such a requirement. 

The TIN Matching service and SSNVS are federal on-line services that 
some private organizations may use voluntarily to verify whether 
federal records show that the name and SSN provided by an individual 
match.[Footnote 37] TIN Matching is a pre-return filing service offered 
by IRS that allows those payers whose income is subject to backup 
withholding, who submit any of six Form 1099 information returns (e.g., 
financial institutions), to match the TIN of the 1099 payee against IRS 
records.[Footnote 38] It is one of several e-service products offered 
by IRS. The goal of TIN Matching is to improve the accuracy of Form 
1099 data and reduce subsequent inappropriate penalties and error 
notices. SSNVS is a service offered by SSA that allows registered users 
(i.e., employers or, in certain instances, their third-party 
representatives) to verify the names and SSNs of employees against SSA 
records.[Footnote 39] The AEITC is outside the scope of SSA's 
responsibilities, and SSNVS is a voluntary service that is currently 
used only to increase the accuracy of wage reporting on Forms W-2. 

IRS and SSA officials identified a number of challenges that the 
agencies and employers may face if the TIN Matching service or SSNVS 
were used to verify AEITC eligibility. 

Accuracy: Both the TIN Matching service and SSNVS are based on SSA 
records and have high rates of accuracy in terms of determining whether 
submitted names and SSNs match. Still, IRS and SSA officials had 
concerns about whether these rates were high enough for purposes of 
verifying AEITC eligibility. IRS officials told us that the TIN 
Matching service is about 98 percent accurate. A December 2006 report 
by the SSA Office of Inspector General sampled more than 2,000 
determinations by SSA's Numident file--the database upon which SSNVS is 
based--and found a name and SSN match accuracy rate of more than 99 
percent.[Footnote 40] Still, SSA officials said that if SSA records 
were used to verify AEITC eligibility, they might want to subject an 
employee's name and SSN to more "routines"--procedures such as 
correcting for transposed numbers that SSA uses to increase the 
likelihood of a match--than is currently done by SSNVS. IRS officials 
told us that when the agency was informally considering a proposal to 
charge a fee for using the TIN Matching service, several IRS officials 
knowledgeable about the database opposed the idea because they did not 
believe it was accurate enough that users should have to pay for it. 

Accuracy concerns, however, do not preclude IRS or SSA from using SSA 
records to make an initial determination about whether individuals who 
may claim credits or benefits have demonstrated that they are entitled 
to them. For example, as previously noted, IRS rejects electronically 
filed tax returns with a name/SSN mismatch and returns them to the 
taxpayer for correction. For paper returns reporting AEITC receipt that 
have a name/SSN mismatch, IRS processes the returns, but since the 
AEITC is reported as tax, it will either offset all or part of any 
refund due or, if no refund is due, require the taxpayer to pay back 
the full AEITC amount. 

Similarly, SSA instructs its claims representatives that if the 
identity of a claimant for Social Security benefits or Supplemental 
Security Income benefits remains questionable because the individual 
has not provided sufficient proof to establish his or her identity, the 
claim will be denied even if other factors of eligibility are 
met.[Footnote 41] 

Additional employer responsibilities: IRS and SSA officials said a 
major concern about employers using either service was whether a name/ 
SSN mismatch would create new responsibilities for employers beyond 
informing the employee and denying the AEITC. IRS officials also 
expressed concern that requiring employers to use the services for 
AEITC would discourage them from promoting the AEITC and perhaps 
encourage them to dissuade employees from seeking it. Under current 
procedures for using the TIN Matching service and SSNVS, employers are 
not required to take any action based on the results they receive. 

IRS officials expressed concern that a June 2006 regulation proposed by 
the Department of Homeland Security could expressly list employer 
receipt of a "no match" letter from SSA as possible evidence that the 
employer knew or should have known that it was employing an individual 
not authorized to work in the United States.[Footnote 42] Under the 
proposed regulation, if the employer fails to take reasonable steps to 
resolve the discrepancy after receiving the letter, the Department of 
Homeland Security may find that the employer had such knowledge and 
assess civil monetary penalties against the employer. 

The proposed Department of Homeland Security regulation describes "safe 
harbor" procedures that the employer can follow in response to the 
letter. Those steps include the employer promptly checking its records 
for clerical errors and obtaining required documentation by working 
with the employee, SSA, and the Department of Homeland Security. If the 
name/SSN mismatch cannot be resolved, the employer would have to choose 
between terminating the employee or facing the risk that the Department 
of Homeland Security may find that the employer knew that the employee 
was not authorized to work and, by continuing to employ the individual, 
violated the law. 

While the proposed Department of Homeland Security regulation covers 
only no match letters, IRS and SSA officials expressed concern that the 
regulation could be expanded to include name/SSN mismatches disclosed 
by the TIN Matching service or SSNVS. Under current law, existing and 
prospective users who are concerned that matching could be used for 
purposes beyond improving the accuracy of Form 1099 data or wage 
reporting can voluntarily cease using, or not start using, the systems. 
IRS and SSA officials noted, however, that employers would no longer 
have this choice if they were required to use one of these services to 
verify an AEITC applicant's SSN. 

In addition, existing Department of Homeland Security guidance for 
employers on the interaction between antidiscrimination laws and legal 
requirements for verifying employment eligibility states that employers 
must treat all employees in the same manner. Employers cannot set 
different employment eligibility verification standards or require that 
different documents be presented by different groups of employees. If 
mandatory verification of AEITC applicants' SSNs created additional 
employer responsibilities under employment eligibility verification 
requirements, this result could be inconsistent with efforts to ensure 
that verification procedures apply to all employees. 

SSA's position on SSNVS is that a name/SSN mismatch does not make any 
statement about an employee's immigration status and should not be a 
basis for taking any adverse action against the employee. SSA officials 
also expressed reservations about SSNVS results being used to terminate 
employees. 

Capacity and User Access: IRS and SSA officials said changes to the 
capacity and user access of the TIN Matching service or SSNVS would 
either be unnecessary or minor if employers used them to verify SSNs of 
employees seeking AEITC, although the officials said their agencies 
might favor creating a new service for this purpose instead. They told 
us that their systems have the capacity to handle the increased volume 
of requests that would result from this expanded use. Because both 
services already are used to verify SSNs, IRS and SSA said any changes 
to how the services are accessed and used would also probably not be 
extensive. 

Both SSNVS and TIN Matching are Web-only services.[Footnote 43] To use 
one of the services, the employer designates one or more employees or 
third-party representatives to register on behalf of the employer. The 
initial registration for both the TIN Matching service and SSNVS is 
handled in a similar way and may take as long as 4 weeks: 

* Registrants go to the agency's Web site and provide information about 
themselves and their employers on a form, which they send 
electronically to the agency. 

* The agency sends the employer of the registrant a unique code. The 
letter directs the employer to provide that code to the registrant. 

* After receiving the code, the registrant can go back to the agency 
Web site and input the code to activate use of the service. 

TIN Matching registrants who do not use it for 6 months must 
reregister, primarily to receive a new password and update any of the 
information provided during the registration. SSNVS requires 
registrants to change their password once a year to keep it from 
expiring, which also requires reregistration. 

IRS and SSA officials told us that the great majority of users of their 
respective services generally report that they are not difficult to use 
for either the registration process or ongoing use. IRS officials said 
users access TIN Matching voluntarily and that some, particularly from 
smaller organizations, appear more likely to find it burdensome than 
users from larger organizations. We found that more than half the 
employers that provided the AEITC in tax years 2002 through 2004 were 
small businesses or self-employed and about one-quarter were tax exempt 
and government entities. In addition, despite IRS's efforts to outreach 
to large employers, fewer than one-fifth were large and midsize 
employers (see table 20 in app. III). IRS officials said TIN Matching 
service users who found the registration process burdensome were 
generally those who reported to IRS that they were not use to filling 
out forms on-line and creating and using passwords. The officials also 
said that some TIN Matching users reported being uncomfortable having 
to provide personal information to register. 

SSA officials said SSNVS is used mostly by larger employers, and a 
relatively small number of them reported that they found the service 
burdensome. SSA officials did say, however, that SSA received about 
89,000 calls through June 2007 from individuals about registration. 
And, an SSA official said that small business representatives with whom 
he has recently spoken expressed frustration with the overall number of 
tasks that the federal government was already requiring them to perform 
and, therefore, might be reluctant to verify SSNs. 

IRS officials also told us that the TIN Matching service is not 
programmed to track users, which likely would be useful to IRS for 
enforcement purposes. Still, IRS officials added, that the service 
could be modified to track employers that used the services for AEITC 
purposes and report on the results. SSNVS tracks its users to determine 
whether the service is being properly used. 

Additional resources: IRS and SSA officials said that if employers were 
required to begin using their respective services to verify the SSNs of 
employees seeking the AEITC, the agencies would need additional 
resources. For example, officials from both agencies cited the need to 
handle questions from new users, particularly in the first year when 
all individuals seeking the AEITC would be required to have their names 
and SSN's matched.[Footnote 44] 

When we told IRS officials that our data showed that about 50,000 
employers had at least one AEITC recipient, they said such an increase 
in the number of registrants could require IRS to increase the number 
of staff available to answer user questions. But they said they could 
not estimate how many more staff would be necessary. 

IRS officials also said they were trying to make the reregistration 
process easier because they received a substantial number of calls from 
users who needed to reregister. This would be important for AEITC- 
related use of TIN Matching because employers with only one or two 
employees seeking the AEITC would only need to use the service once or 
twice a year, making it likely that they would have to reregister. 

SSA officials told us that if the agency was considering or was 
directed to make its SSN records available for the purpose of verifying 
AEITC eligibility, SSA would have to devote additional resources to 
conduct a comprehensive assessment to determine the changes necessary 
to SSNVS to properly achieve this goal, including possibly creating a 
different service for assessing AEITC eligibility and buying a new 
database server to handle the increased volume of users. Again, SSA 
officials could not estimate how many more staff would be necessary. 

New federal legislation: Enactment of federal legislation would be 
needed for employers to begin using the TIN Matching service to verify 
the SSNs of their employees seeking the AEITC.[Footnote 45] In 2000, 
the Department of the Treasury recommended to the Congress that TIN 
verification be expanded to include other payers subject to an IRS 
reporting requirement, such as employers who file Forms W-2.[Footnote 
46] 

It is uncertain whether IRS could require employers to use SSNVS to 
verify the SSNs of employees seeking the AEITC. SSA officials said they 
would need to determine whether SSA's disclosure of SSN data is 
compatible with the reason it collected the information and, if so, 
whether verifying SSNs via SSNVS for purposes of AEITC eligibility is 
consistent with SSA's legal obligations. Both Treasury and SSA 
officials said their agencies would strongly prefer enactment of 
legislation before requiring employer or any verification of SSNs of 
employees seeking the AEITC. 

Employee appeal of mismatch: One difference between using the TIN 
Matching service or SSNVS would occur when an employee claimed that a 
name and SSN mismatch was inaccurate. IRS and SSA officials said 
employees who questioned a SSN mismatch would presumably contact IRS, 
which would send them to SSA to resolve the issue because the TIN 
Matching service is based on SSA records. IRS officials said that, in 
contrast, employees questioning a name and SSN mismatch generated by 
SSNVS would presumably go directly to SSA. 

Agency mission: SSA officials also said that verifying eligibility for 
the AEITC is most appropriate for IRS because it is a tax 
administration issue and is therefore outside the scope of SSA's 
mission. However, it is not unusual for agencies to assist other 
federal agencies in carrying out their mission. SSA officials also said 
that, regardless of which service was used, IRS would have full 
administrative responsibility for overseeing a program for employers to 
verify AEITC applicants' SSNs. 

IRS Does Not Require Submission of Forms W-5 from Employers for 
Employees Seeking the AEITC: 

IRS does not require employers to submit a Form W-5 when an employee 
requests receipt of the AEITC. Several advantages and disadvantages 
exist if IRS creates a Form W-5 database to use in monitoring AEITC 
noncompliance issues. 

IRS could require employers to submit a Form W-5 when an employee 
requests receipt of the AEITC. In turn, IRS could use the Forms W-5 to 
create a database to monitor the AEITC. The database could be used to 
ensure that the SSN provided on the Form W-5 is valid and that it 
matches the individual's name. Doing such a check could have prevented 
more than 100,000 individuals from receiving as much as between $37 
million and $39 million each year in AEITC to which they were 
potentially not entitled because of not meeting the valid SSN 
requirement. Such a database could also allow IRS to know which 
individuals received the AEITC and provide the agency with an 
opportunity to send recipients a notice at the start of the next filing 
season reminding them to file a federal tax return. A reminder to file 
notice could likely reduce noncompliance for up to about 200,000 
individuals who received between $42 million and $50 million each year 
in AEITC without filing a federal tax return. Similarly, IRS officials 
could use a W-5 database to verify other AEITC requirements, such as 
ensuring that each recipient has only one Form W-5 in effect at a time. 
This check could reduce the probability that individuals would receive 
more than the yearly AEITC maximum. 

While acknowledging that potential advantages exist to developing and 
maintaining a Form W-5 database, IRS officials said that the 
disadvantages could outweigh these and any other advantages. Although 
IRS officials said it was too early in the proposal process to 
calculate the database's potential costs and subsequent return on 
investment, they said it very likely would be substantially lower than 
the return on investment for either existing or anticipated future 
noncompliance programs. For example, IRS estimates the current return 
on investment for EITC Examination noncompliance is between $17 to $1 
and $19 to $1. Although these amounts only include labor and do not 
include overhead such as facilities, equipment, and supplies, officials 
felt confident that the EITC return on investment would far exceed that 
for the AEITC. Their opinion was largely based on the few dollars 
involved with the AEITC, especially compared to other noncompliance 
programs. 

In addition, IRS officials expressed concerns that employers would not 
submit Forms W-5 to IRS. Officials raised an analogy between this 
proposal and the prior Questionable W-4 program.[Footnote 47] As we 
reported in 2003, about 75 percent of the large employers with 1,000 or 
more employees in IRS's Large and Medium-Size Business and Small 
Business/Self Employed divisions who filed tax returns in tax year 2001 
did not send IRS any questionable Forms W-4.[Footnote 48] After our 
report, IRS discontinued the Questionable W-4 program. Additionally, 
officials noted that requiring employers to submit Forms W-5 may 
discourage them from participating because if the employer was notified 
that the SSN on a Form W-5 was invalid or that it did not match the 
employee's name, employers would likely have a responsibility to 
discuss the matter with the employee, creating yet another new burden 
employers would not want to accept. Finally, employers may not 
participate because if the employee left the employer during the year, 
the employer would again have to contact IRS so the Form W-5 database 
could be updated. 

Options to Reduce AEITC Noncompliance May Be Cost Effective: 

Although we do not know how successful the various options we have 
identified for improving AEITC compliance may be if implemented or what 
the full cost of implementation would be, IRS may be able to achieve a 
return on investment somewhat comparable to that for EITC examinations. 
We found an average of about $94 million a year in AEITC noncompliance 
for recipients in tax years 2002 through 2004.[Footnote 49] If a 
compliance effort could reduce AEITC noncompliance by one-quarter, that 
is, $24 million per year, IRS could spend about $1.3 million each year 
to do so and achieve a $19 to $1 return on investment.[Footnote 50] 
Alternatively, IRS estimated that it would cost about $533,000 to send 
soft notices to 300,000 taxpayers during the Dependent Database test. 
If IRS were to test sending soft notices for AEITC and it cost IRS 
about the same amount to send notices to 300,000 noncompliant AEITC 
recipients, IRS would only need to reduce AEITC noncompliance by about 
11 percent (about $10 million) to achieve a $19 to $1 return on 
investment. 

Eligible Taxpayers Could Receive Full EITC Benefits If the Advance 
Option Were Discontinued: 

If the advance option were discontinued, eligible AEITC recipients 
could still receive the full benefits of the EITC as a lump sum after 
filing their tax return. In addition, improper AEITC payments to 
ineligible or noncompliant individuals would be eliminated. The exact 
amount of revenue that could be saved is not known. However, in 
determining an amount, IRS officials said they would consider the 
following: the amount of AEITC disbursed compared with the amount shown 
on filed tax returns, the cost of administering the AEITC (e.g., forms 
and publications, processing, compliance activities), and any amount 
currently recovered through compliance activities. 

Conclusions: 

The AEITC has never achieved a significant participation rate, the 
amount recipients received in the period we reviewed was quite low, and 
noncompliance was high. However, policymakers may judge that the goal 
of providing funds to low-income workers during the year, as opposed to 
a lump sum that they could get as part of the EITC when filing their 
taxes, remains important and should continue to be allowed. If so, IRS 
needs to pursue potentially cost-effective measures to address AEITC 
noncompliance. 

Each of the three options we identified for improving AEITC compliance-
-using soft notices, having up-front verification of AEITC applicants' 
SSNs by employers, or requiring employers to submit copies of the 
Form(s) W-5 and creating a database to monitor AEITC--appears to have 
potential to improve compliance, but their full benefits and costs need 
to be evaluated and, if possible, tested. Soft notices have improved 
compliance in other tax programs but they could be somewhat less 
effective in improving AEITC compliance, in part, due to its high 
turnover rate. The TIN Matching service and SSNVS have the potential to 
reduce AEITC noncompliance by enabling employers to verify workers' 
SSNs before providing them the AEITC. Differences exist between the two 
services, but either could likely be used for AEITC SSN verification. 
Both services are based on SSA records that are already deemed accurate 
enough such that SSA and IRS make decisions based on them to disallow 
certain exemptions and credits until eligibility has been properly 
demonstrated. Significant concerns exist, however, such as the need for 
legislation authorizing either the use of TIN Matching or SSNVS for 
AEITC. This and other issues would need to be further explored as the 
costs and benefits of employers verifying employees' SSNs are fully 
identified. If IRS required employers to submit copies of the Form W-5 
when an employee requests the AEITC, IRS could create a database to 
better monitor and address all three of the noncompliance problems we 
analyzed. However, imposing additional responsibilities on employers 
for both the SSN verification option and the Form W-5 database option 
have the potential to adversely affect the AEITC's already low 
participation rate if employers avoid providing the AEITC due to 
increased responsibilities on their part. 

Due to the relatively small size of the AEITC overall, combined with 
the low dollar amounts per taxpayer, IRS officials are concerned that 
addressing AEITC noncompliance may provide less return on IRS's 
enforcement efforts than would addressing other noncompliance issues. 
However, IRS may be able to achieve returns on AEITC enforcement that 
would not be significantly out of line with returns on other 
enforcement work. For example, it cost IRS about $533,000 to send soft 
notices to 300,000 taxpayers in the Dependent Database test. If IRS 
were to test sending soft notices for AEITC and it cost IRS about the 
same amount to send notices to 300,000 noncompliant AEITC recipients, 
IRS would need to reduce noncompliance by about 11 percent (about $10 
million) to achieve a $19 to $1 return on investment. 

Recommendations for Executive Action: 

The Acting Commissioner of Internal Revenue should analyze whether any 
of the following options could cost effectively and significantly 
reduce AEITC noncompliance: 

* sending potentially noncompliant AEITC recipients soft notices, such 
as to nonfilers whose Forms W-2 show that they received AEITC and 
filers who misreported the amount they received or whose SSN and name 
do not match; 

* requiring employers to verify the SSN of employees seeking AEITC; or: 

* requiring employers to submit Form W-5 to IRS and IRS creating and 
maintaining a database for these forms. 

To better identify the costs and implementation issues as well as the 
likelihood for these or other options to reduce AEITC noncompliance, 
where practical, the Acting Commissioner of Internal Revenue should 
test these options to make a more fully informed judgment about whether 
any would be worthwhile. 

If the Acting Commissioner of Internal Revenue determines that none of 
these options would be cost effective and no other remedies are viable, 
then the Treasury Secretary should inform the Congress of this and 
provide Treasury's opinion about whether the AEITC should be retained. 

Agency Comments: 

The Acting Commissioner of Internal Revenue provided written comments 
in a July 18, 2007 letter. He agreed with our recommendation and 
outlined the actions IRS would take to address that recommendation, 
including conducting further analyses and possible testing of proposed 
options for reducing AEITC noncompliance. He also stated that IRS will 
conduct its cost-benefit analyses in conjunction with a congressional 
requirement to study the impact of expanding eligibility of the AEITC 
to all EITC recipients. We also provided a draft of this report to the 
Department of the Treasury and SSA and incorporated technical comments 
where appropriate. SSA emphasized that verifying eligibility for the 
AEITC is most appropriate for IRS because it is a tax administration 
issue and therefore outside the scope of SSA's mission. 

As agreed with your offices, unless you publicly announce its contents 
earlier, we plan no further distribution of this report until 30 days 
after its date. At that time, we will send copies of this report to the 
Secretary of the Treasury, the Commissioner of the Social Security 
Administration, the Acting Commissioner of Internal Revenue, 
appropriate Congressional committees, and other interested parties. 
This report is available at no charge on GAO's web site at [hyperlink, 
http://www.gao.gov]. 

If you or your staffs have any questions, please contact me at (202) 
512-9110 or brostekm@gao.gov. 

Signed by: 

Michael Brostek: 
Director, Tax Issues: 
Strategic Issues Team: 

[End of section] 

Appendix I: IRS's Implementation of Recommendations from GAO's 1992 
Report on the Advance Earned Income Tax Credit: 

After analyzing IRS's responses to our recommendations in our 1992 
Advance Earned Income Tax Credit (AEITC) report, we determined that IRS 
has implemented five of the six recommendations to the Commissioner of 
Internal Revenue from our 1992 report on the AEITC and partially 
implemented the remaining one (see table 8). 

Table 8: IRS Implementation of Recommendations from GAO's 1992 Report 
on the Advance Earned Income Tax Credit: 

Number: 1; 
Recommendation: Include information on AEITC in employee outreach 
materials and programs; 
Status: Implemented; 
IRS implementation: Developed publicity materials. 

Number: 2; 
Recommendation: Notify taxpayers who receive the Earned Income Tax 
Credit (EITC) about the advance option; 
Status: Implemented; 
IRS implementation: Revised Form W-2 and Notice 797 to contain 
information on how to apply for the AEITC and developed radio and 
television announcements on the AEITC. 

Number: 3; 
Recommendation: Encourage employers to notify employees who have no 
income tax withheld of the advance option; 
Status: Implemented; 
IRS implementation: Added text to the face of the Form W- 4 
instructions advising employees to consider the AEITC. 

Number: 4; 
Recommendation: Clarify instructions on AEITC in "Circular E, 
Employer's Tax Guide."; 
Status: Implemented; 
IRS implementation: Clarified employers' responsibilities and 
liabilities in advancing the EITC in the 1993 "Circular E, Employer's 
Tax Guide.". 

Number: 5; 
Recommendation: Send, to individuals who do not file tax returns, a 
notice explaining their requirements to file; 
Status: Partially Implemented; 
IRS implementation: Included information on the AEITC in a reminder to 
file notice until 1997 and added a separate AEITC box on Form W-2. 

Number: 6; 
Recommendation: Explore ways to identify those individuals who receive 
the credit advance, but do not report it, so as to prevent them from 
receiving the credit a second time; 
Status: Implemented; 
IRS implementation: Added a line on the Form 1040 for taxpayers to 
report AEITC received and redesigned the Form W-2. 

Source: GAO analysis of IRS data. 

[End of table] 

Our first recommendation was for the Commissioner of Internal Revenue 
to include information on AEITC in employee outreach materials and 
programs. IRS implemented this recommendation primarily by developing 
publicity materials (i.e., grocery bags, milk carton art, brochures, 
posters) and distributing them to the public. 

For the second recommendation, IRS stated that it did not have the 
approximately $2 million in funding that the agency said would have 
been required to notify all taxpayers who receive the EITC, but did not 
elect the advance option. Instead, IRS took other actions including 
revising the Form W-2 in 1992 to contain information on how to apply 
for the AEITC and IRS Notice 797, "Possible Federal Tax Refund Due to 
the Earned Income Credit," to include information on how to apply for 
the AEITC. 

For our third recommendation, IRS noted that it could encourage 
employers to make such notifications, but that there are no statutory 
sanctions on employers who fail to do so. Beginning in 1992, IRS placed 
text on the face of the Form W-4 instructions advising employees to 
consider filing a Form W-5 with their employer to obtain the advance 
through lower withholding. 

For our fourth recommendation to clarify instructions on the AEITC in 
"Circular E, Employer's Tax Guide" IRS did so through the inclusion of 
new examples explaining to employers how they should make advance 
payments to employees and how employers can report these amounts. 

Our fifth recommendation was for the IRS Commissioner to send 
individuals who received the AEITC and do not file tax returns a notice 
explaining the requirement to file. IRS partially implemented this 
recommendation by including information on advance payment in a 
reminder to file notice and adding a separate AEITC box on Form W-2. 
IRS did not track the number of AEITC nonfilers who received the 
notice. The reminder to file notice was only sent until 1997 and IRS 
officials were uncertain why that notice was discontinued. 

The last recommendation was for exploring ways to identify those 
individuals who receive the credit in advance but do not report it. IRS 
pointed out that its systems were not geared to detecting unreported 
AEITC payments at the time the returns are processed and the best 
approach to preventing noncompliance by AEITC payment recipients is a 
proactive one that recognizes the filing of correct returns. IRS 
implemented this recommendation by providing a separate line on Form 
1040 on which to report AEITC payments and redesigning the Form W-2, 
for tax year 1993, which it believed would increase the accuracy of the 
AEITC payment information reported on Form W-2. Our work in this report 
demonstrates a continuing need to explore additional compliance 
initiatives aimed at those who receive the AEITC, but do not report it 
on their tax return. 

[End of section] 

Appendix II: Scope and Methodology: 

To answer the first and second objectives: 

* how many individuals received the Advance Earned Income Tax Credit 
(AEITC) compared with the Earned Income Tax Credit (EITC) and how much 
did they receive in tax years 2002 through 2004; what actions, if any, 
have been taken to increase use since 1992; and what is the potential 
for significant increases in the future; and: 

* what is the extent of noncompliance, if any, associated with the 
AEITC; 

we obtained a data file of all Forms W-2, "Wage and Tax Statement," for 
tax years 1999-2004 indicating AEITC payments as shown by an amount 
greater than $0 in box 9 of the Form W-2 from the Internal Revenue 
Service (IRS).[Footnote 51] We used these tax years because they were 
the most current available at the time we started our review. The Form 
W-2 identified key information, including the AEITC recipient's name, 
address, Social Security number (SSN), and amount of AEITC dollars 
paid, as well as the employer's name and address. 

To determine the number of individuals who received the AEITC, we used 
Forms W-2 instead of tax returns, which IRS has historically used to 
estimate AEITC use. We used this alternate approach because we believe 
the Forms W-2 provide results that are more accurate and 
complete.[Footnote 52] For example, using Forms W-2 would include in 
the population of AEITC recipients those who received AEITC, but did 
not file a return, and those who filed a return, but did not report any 
AEITC. Using tax returns would not capture these individuals or related 
noncompliance issues. In addition, using tax returns counts instances 
where both spouses receive the AEITC and file jointly on one return as 
opposed to two individuals. IRS's Research, Analysis, and Statistics 
and EITC program office officials agreed with our methodology. 

We performed data reliability tests on the data file to determine 
whether the data were sufficiently reliable for our intended purposes. 
We did this testing, in part, by conducting preliminary analyses, which 
identified certain data irregularities or anomalies. We identified two 
noteworthy anomalies in the data file: (1) excessive AEITC dollar 
amounts and (2) invalid AEITC recipient SSNs. First, many Forms W-2 
showed that employees received amounts over the allowable limits. A few 
even showed individuals each receiving about $1 million in AEITC-- 
amounts clearly above AEITC legal limits and which IRS officials said 
would be improbable, potentially resulting from transcription errors. 
Second, we also found some instances where the SSN and/or name on the 
Form W-2 were invalid, which means that the number was never issued by 
SSA or that the name and number on the Form W-2 did not match the 
listed name for that same SSN in official records maintained by IRS. We 
compared the number and name information on the Form W-2 to the 
National Account Profile to evaluate the validity of that information 
and to identify any possible subsequent corrections.[Footnote 53] 

To address these data anomalies, we separated the Form W-2 file into 
four subpopulations using the following three criteria: whether (1) the 
SSN on the Form W-2 was valid, according to Data Master File (DM-1); 
(2) the SSN and the recipient's name on the Form W-2 matched, according 
to DM-1; and (3) the amount of AEITC received was in excess of the 
yearly maximum.[Footnote 54] Each of the subpopulations had a unique 
profile, as follows:[Footnote 55] 

1. Valid subpopulation: This group of Forms W-2 represents all 
individuals (1) that had a valid SSN, meaning that it was a number 
issued by SSA, (2) whose name matches the SSN, and (3) that had an 
AEITC amount within the yearly maximum. More than 75 percent of the 
Forms W-2 on average during tax years 2002 through 2004 were in this 
subpopulation. 

2. Invalid name subpopulation: This group of Forms W-2 represents all 
individuals that had (1) a valid SSN (2) a SSN that did not match the 
individual's name and (3) the AEITC amount was within the yearly 
maximum. About 17 percent of the Forms W-2 fell in this subpopulation 
for each of the 3 years we reviewed. 

3. Invalid number subpopulation: This group of Forms W-2 represents all 
Forms W-2 that had an invalid SSN and an AEITC amount that was within 
the yearly maximum. About 7 percent of the Forms W-2 during tax years 
2002 through 2004 were in this population.[Footnote 56] 

4. Dollar limit subpopulation: This group of Forms W-2 represents all 
instances where the AEITC amount was above the yearly maximum, 
regardless of whether the SSN was invalid or if the individual's name 
matched the SSN. This represented less than 1 percent on average of all 
Forms W-2 in each of the 3 years we reviewed. Because IRS officials 
told us these data were likely erroneous, we excluded it from most of 
our analyses, and IRS officials agreed. 

Table 9: Number of Forms W-2, Number of Individuals, and the Dollars 
for Each Subpopulation, Tax Years 2002 through 2004: 

Valid subpopulation; (percentage); 
Tax year 2002: Number of Forms W-2: 488,007: (74%); 
Tax year 2002: Number of individuals: 418,774: (78%); 
Tax year 2002: Dollars: $107,230,369: (57%); 
Tax year 2003: Number of Forms W-2: 463,285: (76%); 
Tax year 2003: Number of individuals: 398,526: (79%); 
Tax year 2003: Dollars: $106,075,016:(70%); 
Tax year 2004: Number of Forms W-2: 469,234:( 76%); 
Tax year 2004: Number of individuals: 398,335: (78%); 
Tax year 2004: Dollars: $111,098,330: (68%). 

Invalid name subpopulation:(percentage); 
Tax year 2002: Number of Forms W-2: 114,819: (17%); 
Tax year 2002: Number of individuals: 80,524: (15%); 
Tax year 2002: Dollars: $25,734,033: (14%); 
Tax year 2003: Number of Forms W-2: 101,646: (17%); 
Tax year 2003: Number of individuals: 74,234:(15%); 
Tax year 2003: Dollars: $25,480,158: (17%); 
Tax year 2004: Number of Forms W-2: 101,140: (16%); 
Tax year 2004: Number of individuals: 77,747: (15%); 
Tax year 2004: Dollars: $26,819,235:(16%). 

Invalid number subpopulation:(percentage); 
Tax year 2002: Number of Forms W-2: 49,432: (8%); 
Tax year 2002: Number of individuals: 32,501: (6%); 
Tax year 2002: Dollars: $11,582,604: (6%); 
Tax year 2003: Number of Forms W-2: 44,244: (7%); 
Tax year 2003: Number of individuals: 30,622: (6%); 
Tax year 2003: Dollars: $11,650,727: (8%); 
Tax year 2004: Number of Forms W-2: 43,696: (7%); 
Tax year 2004: Number of individuals: 31,875: (6%); 
Tax year 2004: Dollars: 12,166,702: (7%).

Dollar limit subpopulation:(percentage); 
Tax year 2002: Number of Forms W-2: 6,408:(1%); 
Tax year 2002: Number of individuals: 6,360:(2%); 
Tax year 2002: Dollars: $43,703,747:(23%); 
Tax year 2003: Number of Forms W-2: 2,692:(less than 1%); 
Tax year 2003: Number of individuals: 2,677:(1%); 
Tax year 2003: Dollars: $7,486,726:(5%); 
Tax year 2004: Number of Forms W-2: 2,768:(less than 1%); 
Tax year 2004: Number of individuals: 2,757:(1%); 
Tax year 2004: Dollars: $12,917,134:(8%). 

Total; 
Tax year 2002: Number of Forms W-2: 658,666; 
Tax year 2002: Number of individuals: 538,159; 
Tax year 2002: Dollars: $188,250,753; 
Tax year 2003: Number of Forms W-2: 611,867; 
Tax year 2003: Number of individuals: 506,059; 
Tax year 2003: Dollars: $150,692,627; 
Tax year 2004: Number of Forms W-2: 616,838; 
Tax year 2004: Number of individuals: 510,714; 
Tax year 2004: Dollars: $163,001,401. 

Source: GAO analysis of IRS data. 

Note: Percentages may not add due to rounding. 

[End of table] 

We conducted additional data reliability tests for each of the 
databases we used to obtain information about the AEITC, including 
IRS's Individual Returns Transaction File, for return and filing 
information, which came from the Compliance Data Warehouse; National 
Account Profile/DM-1, for IRS's SSN and name reference information, 
which also came from the Compliance Data Warehouse; Automated 
Underreporter (AUR), for IRS's third-party information return data; and 
Taxpayer Identification Number (TIN) Matching and SSA's Social Security 
Number Verification System (SSNVS), for alternative SSN and name 
reference information used by employers.[Footnote 57] After completing 
our data reliability assessments, we determined the AEITC data to be 
sufficiently reliable for analysis and our reporting objectives. 

We also developed a comprehensive analysis plan that included our 
researchable issues, planned analysis, data sources, and limitations. 
We shared our plan with IRS and others and incorporated their feedback. 
Because IRS's workload precluded them from providing information 
related to employers/AEITC payers within our time frames, we were able 
to conduct only limited analyses of employers who paid AEITC. 

Using the analysis plan for each subpopulation, we conducted multiple 
analyses to develop relevant demographic, characteristic, and 
compliance data. Because each population had different criteria, 
certain characteristics or compliance data could not be developed or 
compared across the subpopulations. For example, the only 
characteristics data that could be developed for the invalid SSN 
subpopulation came from the Form W-2 (e.g., amount of AEITC, geographic 
location) because it is the only available source. Similarly, tax 
return data, such as filing status, was not available for those who did 
not file a tax return. 

All data pertaining to filed tax returns came either from returns that 
reported receipt of the AEITC on the appropriate line or from a 
"constructed tax return," which IRS officials created using the SSN on 
the Form W-2 and matching it to an SSN in the primary, secondary, or 
dependent position on a filed return. The location of the SSN in one of 
these positions is relevant due to the way IRS manages its data files. 
There could be instances when a tax return was filed but it was not 
detected using our methodology. For example, a taxpayer's SSN on the 
Form W-2 might have been incorrect and the taxpayer reported the 
correct number on the tax return (Form 1040). 

To report on data pertaining to the EITC, we relied on published EITC 
data provided by IRS research and program office officials, including 
the EITC database, the EITC Database Year to Year Comparison Report, 
and EITC Fact Sheet.[Footnote 58] When possible, our analysis compares 
individuals who received the AEITC with individuals who received the 
EITC. IRS defines EITC recipients by the number of federal tax returns 
that received EITC. In addition, EITC data are based upon the primary 
TIN of all taxpayers who received an amount of EITC. We determined the 
number of individuals who receive the AEITC based on the number of 
Forms W-2 reporting AEITC per unique SSN. IRS officials agreed that 
even though these populations are not identical, it is reasonable to 
make a comparison between them. 

We frequently consulted with IRS officials on the data and our 
analyses; they generally agreed with both the approach and the accuracy 
of the results. For the analyses that IRS conducted, we agreed with 
both the approach and the accuracy of the results. 

In addition, we reviewed legislative and IRS administrative changes to 
the AEITC since 1992 and discussed them with IRS and other officials, 
including Department of the Treasury officials. We reviewed reports on 
IRS's implementation of some of our prior recommendations pertaining to 
the EITC and discussed them with IRS officials, including the National 
EITC Director. We also coordinated this work with the Treasury 
Inspector General for Tax Administration. 

Finally, we also identified and interviewed 11 individuals we 
determined to be experts to provide us a fuller understanding on the 
potential for significant increases or improvement to AEITC use and 
noncompliance and included academics, researchers, practitioners, and 
individuals representing the areas of tax policy, low-income individual 
issues, and compliance issues.[Footnote 59] We chose these individuals 
based on our knowledge of their areas of expertise and our research 
that indicated they were knowledgeable about the EITC. 

To address the third objective, how well do IRS's procedures address 
any areas of noncompliance, we examined portions of IRS's Internal 
Revenue Manual and interviewed IRS Wage and Investment and Small 
Business/Self-Employed division officials to determine procedures for 
processing returns that reported receipt of the AEITC. We examined how 
IRS's enforcement procedures and operations, including Submission 
Processing, AUR, Nonfiler, Collection, Examination, Criminal 
Investigation, and Taxpayer Advocate, address certain kinds of 
potential noncompliance. 

We explored various options to improve AEITC compliance. This involved 
conducting literature searches and interviews with IRS and SSA 
officials. We reviewed and discussed the results of soft notice tests 
with IRS officials, including the National EITC Director, and discussed 
the applicability of soft notices for addressing AEITC noncompliance. 
We also reviewed and analyzed documents and reports about IRS, SSA, and 
Department of Homeland Security databases about whether they could be 
used by employers to verify the SSN of an employee seeking the AEITC 
before the employer begins paying it. We also interviewed knowledgeable 
officials at IRS and SSA about the advantages and disadvantages of such 
systems when considering the AEITC. Further, we interviewed IRS 
officials from various offices, such as EITC Program, Modernization and 
Information Technology Services, and Stakeholders, Partnership, 
Education and Communication, about the advantages and disadvantages of 
creating a database for the Forms W-5. It was not within the scope of 
our work to fully evaluate the potential cost and benefits of these 
options for reducing noncompliance. 

We also reviewed prior GAO, IRS, Treasury Inspector General for Tax 
Administration, and other reports on the AEITC and EITC. 

We conducted our work primarily in Washington, D.C., and Atlanta, Ga., 
from December 2005 through July 2007 in accordance with generally 
accepted government auditing standards. 

[End of section] 

Appendix III: Demographic Characteristics of Advance Earned Income Tax 
Credit Recipients and Their Employers: 

We identified basic demographic characteristics of Advance Earned 
Income Tax Credit (AEITC) recipients and their employers in tax years 
2002 through 2004. Specifically, we identified the following 11 
characteristics: (1) number of Forms W-2 received; (2) average amount 
of AEITC received by consecutive recipients; (3) average adjusted gross 
income (AGI) for AEITC recipients; (4) average wages; (5) filing 
status; (6) age; (7) gender; (8) number of qualifying children; (9) 
filing method; (10) geographic location of AEITC recipients; and (11) 
employer size, number of employees, and number of Forms W-2 with AEITC 
issued to employees. Each of the characteristics represents an analysis 
and provides additional objective information about AEITC recipients 
not previously discussed. Where possible, we compared AEITC recipients 
to Earned Income Tax Credit (EITC) recipients. Further analyses may 
provide information to better target IRS enforcement efforts. For 
example, IRS's information on EITC associated with gender differs from 
our AEITC results in table 17 on gender. 

The characteristics are organized by a declining AEITC population size. 
For example, table 10, "Number of Forms W-2 Received by Subpopulation, 
Tax Years 2002 through 2004," includes the valid, the invalid name, and 
invalid number subpopulations, as described in the scope and 
methodology (see app. II), while table 15, "Filing Status of AEITC and 
EITC Recipients, Number and Percentage, Tax Years 2002 through 2004," 
includes only the valid and invalid name subpopulations. The invalid 
number subpopulation was not included in the table about filing status 
because that subpopulation contains only Form W-2 data and not tax 
return data. Thus, information such as filing status, which comes from 
the tax return, is not available. 

Number of Forms W-2 with AEITC that each individual received: Most 
individuals who received the AEITC only had one Form W-2 reporting its 
receipt (see table 10). Having more than one Form W-2 does not 
necessarily indicate noncompliance because an individual may have more 
than one job during the year and receive the AEITC from more than one 
employer. The data did not enable us to analyze whether any of these 
individuals had more than one Form W-5 in effect at one time. 
Presently, IRS's administrative procedures do not enable it to identify 
whether the taxpayer has more than one Form W-5 in effect at one time. 

Table 10: Number of Forms W-2 Received by Subpopulation, Tax Years 2002 
through 2004: 

[See PDF for image]

Source: GAO analysis of IRS data. 

Note: This table includes the valid, invalid name, and invalid number 
subpopulations. About 20 percent of the combined invalid name and 
invalid number subpopulations were noncompliant (see table 4). 

[End of table] 

Average amount of AEITC received by consecutive recipients: About 
98,000 individuals received the AEITC consecutively during tax years 
2002 through 2004. These 98,000 individuals received a higher average 
amount of AEITC than the entire AEITC population (see fig. 6). 

Figure 6: Average Amount of AEITC Received by Consecutive and AEITC 
Recipients, Tax Years 2002 through 2004: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

Note: This figure includes the valid, invalid name, and invalid number 
subpopulations. 

[End of figure] 

Adjusted gross income (AGI) for AEITC and EITC recipients: The maximum 
amount of AGI a taxpayer could have in tax years 2002 through 2004 and 
receive the AEITC and/or EITC was $34,178, $34,692 and $35,458, 
respectively. As noted in tables 11, 12, and 13, most individuals who 
received the AEITC and filed a tax return reported an AGI of $1-- 
$20,000. Some taxpayers had an AGI above the allowable limits. However, 
because an individual's personal circumstance may have changed during 
the year, for example the individual may have gotten a higher paying 
job, as long as the same amount of AEITC received as shown on the Form 
W-2 was reported on the tax return, AGI outside the limit for AEITC 
recipients is permissible and the taxpayer is considered compliant. By 
reporting the correct amount on the tax return, the AEITC would 
increase the tax due or reduce any refund. 

Table 11: AGI for AEITC and EITC Recipients, Number, Average, and Sum, 
for Tax Year 2002: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

[A] Negative dollar amounts are possible with an AGI less than or equal 
to $0 because these returns likely have eligible deductions (such as 
alimony payments) that exceed the AGI amount, thus resulting in a 
negative AGI. This table includes the valid and invalid name 
subpopulations. 

[End of figure] 

Table 12: AGI for AEITC and EITC Recipients, Number, Average, and Sum, 
for Tax Year 2003: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

[A] Negative dollar amounts are possible with an AGI less than or equal 
to $0 because these returns likely have eligible deductions (such as 
alimony payments) that exceed the AGI amount, thus resulting in a 
negative AGI. This table includes the valid and invalid name 
subpopulations. 

[End of table] 

Table 13: AGI for AEITC and EITC Recipients, Number, Average, and Sum, 
for Tax Year 2004: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

[A] Negative dollar amounts are possible with an AGI less than or equal 
to $0 because these returns likely have eligible deductions (such as 
alimony payments) that exceed the AGI amount, thus resulting in a 
negative AGI. This table includes the valid and invalid name 
subpopulations. 

[End of table] 

Wages for AEITC and EITC recipients: The yearly wage limits for AEITC 
and EITC recipients were $34,178 for tax year 2002, $34,692 for tax 
year 2003, and $35,458 for tax year 2004. The average wages for AEITC 
recipients in the valid subpopulation were about $18,000, while they 
were about $47,000 for the invalid name subpopulation.[Footnote 60] 
This compares with about $13,000 for EITC recipients (see table 14). 
Some wages are outside the allowable limits. However, because an 
individual's personal circumstance may have changed during the year, 
for example, the individual may have gotten a higher paying job, as 
long as the same amount of AEITC received as shown on the Form W-2 was 
reported on the tax return, wages outside the limit are permissible and 
the taxpayer is considered compliant. By reporting the correct amount 
on the tax return, the AEITC would increase the tax due or reduce any 
refund. 

Table 14: Wages Reported for AEITC and EITC Recipients, Number, Wage 
Amount, and Average Dollars Received for Tax: 

Tax year: 2002; 
AEITC: Valid subpopulation: Number: 283,316; 
AEITC: Valid subpopulation: Wage amount: $4,926,169,548; 
AEITC: Valid subpopulation: Average: $17,388; 
AEITC: Invalid name subpopulation: Number: 59,525; 
AEITC: Invalid name subpopulation: Wage amount: $2,696,220,392; 
AEITC: Invalid name subpopulation: Average: $45,296; 
EITC: Number: 20,918,823; 
EITC: Wage amount: $273,167,081,421; 
EITC: Average: $13,058. 

Tax year: 2003; 
AEITC: Valid subpopulation: Number: 265,966; 
AEITC: Valid : Wage amount: $4,829,307,231; 
AEITC: Valid subpopulation: Average: $18,158; 
AEITC: Invalid name subpopulation: Number: 54,622; 
AEITC: Invalid name subpopulation: Wage amount: $2,545,765,228; 
AEITC: Invalid name subpopulation: Average: $46,607; 
EITC: Number: 21,431,377; 
EITC: Wage amount: $280,828,962,882; 
EITC: Average: $13,104. 

Tax year: 2004; 
AEITC: Valid subpopulation: Number: 254,003; 
AEITC: Valid subpopulation: Wage amount: $4,409,967,145; 
AEITC: Valid subpopulation: Average: $17,362; 
AEITC: Invalid name subpopulation: Number: 55,721; 
AEITC: Invalid name subpopulation: Wage amount: $2,759,417,493; 
AEITC: Invalid name subpopulation: Average: $49,522; 
EITC: Number: 21,721,218; 
EITC: Wage amount: $290,119,218,681; 
EITC: Average: $13,356. 

Source: GAO analysis of IRS data. 

Notes: This table includes the valid and invalid name subpopulations. 

[End of table] 

Filing status of AEITC and EITC recipients: As shown in table 15, about 
half of AEITC recipients in the valid subpopulation and most 
individuals who received the EITC used the Head of Household filing 
status. This compares to AEITC recipients in the invalid name 
subpopulation who most frequently used the Married Filing Jointly 
filing status. About 2 percent of the tax returns that reported 
receiving AEITC used the Married Filing Separate filing status, which 
is not allowed. However, an individual's personal circumstance may have 
changed during the year, for example the individual may have separated 
from their spouse. As long as the same amount of AEITC received as 
shown on the Form W-2 was reported on the tax return, this situation is 
permissible and the taxpayer is considered compliant. By reporting the 
correct amount on the tax return, the AEITC would increase the tax due 
or reduce any refund. 

Table 15: Filing Status of AEITC and EITC Recipients, Number and 
Percentage, Tax Years 2002 through 2004: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

Notes: EITC totals are the number of taxpayers who received EITC. N/A 
means not applicable. This table includes the valid and invalid name 
subpopulations. 

[End of table] 

Age of AEITC and EITC recipients: Most individuals who received the 
AEITC, as well as most EITC recipients, were between the ages of 26 and 
64 (see table 16). IRS officials noted that recipients whose age fell 
into the 'Over 100' category are the result of a probable error, such 
as a transcription error. Alternatively, an AEITC recipient in this 
category may have used a SSN of a deceased individual. 

Table 16: Age of AEITC and EITC Recipients, Tax Years 2002 through 
2004: 

[See PDF for image]

Source: GAO analysis of IRS data. 

Notes: This table includes the valid and invalid name subpopulations. 
IRS calculated age by subtracting the birth year from the analysis 
year. Neither IRS nor GAO could identify why the tax years 2002 and 
2003 AEITC valid Forms W-2 SSN total is one below the total number of 
individuals who received a Form W-2 reporting AEITC. 

[End of table] 

Gender of AEITC and EITC recipients: More males than females received 
the AEITC during tax years 2002 through 2004. In contrast, more females 
than males received the EITC during this same period (see table 17). 

Table 17: Gender of AEITC and EITC Recipients, Number and Percentage, 
Tax Years 2002 through 2004: 

[See PDF for image]

Source: GAO analysis of IRS data. 

Notes: This table includes the valid and invalid name subpopulations. 
We reported gender information only for individuals who filed a federal 
tax return. Percentages may not add due to rounding. 

[End of table] 

Number of qualifying children for AEITC and EITC recipients: About half 
of the individuals who reported receiving the AEITC had two qualifying 
children. Results were similar for EITC recipients. However, about 8 
percent of this population did not report having any qualifying 
children, which is not allowed (see table 18). However, an individual's 
personal circumstance may have changed during the year, for example, 
the individual may have separated from their spouse or divorced. As 
long as the same amount of AEITC received as shown on the Form W-2 was 
reported on the tax return, the taxpayer is considered compliant. By 
reporting the correct amount on the tax return, the AEITC would 
increase the tax due or reduce any refund. In contrast, most 
individuals who received the EITC also had two qualifying children. 
There is no qualifying child requirement to receive the EITC. 

Table 18: Number of Qualifying Children for AEITC and EITC Recipients, 
Number and Percentage, Tax Years 2002 through 2004: 

Number of taxpayers with no qualifying children (percentage);
Tax year 2002: AEITC: 7,946: (8%); 
Tax year 2002: EITC: 3,765,072: (18%); 
Tax year 2003: AEITC: 7,456: (7%); 
Tax year 2003: EITC: 4,022,684: (19%); 
Tax year 2004: AEITC: 7,786: (8%); 
Tax year 2004: EITC: 4,181,382: (19%). 

Number of taxpayers with 1 qualifying child (percentage); 
Tax year 2002: AEITC: 43,066: (41%); 
Tax year 2002: EITC: 8,198,533: (39%); 
Tax year 2003: AEITC: 43,129: (41%); 
Tax year 2003: EITC: 8,269,050: (39%); 
Tax year 2004: AEITC: 40,379: (40%); 
Tax year 2004: EITC: 8,354,539: (38%). 

Number of taxpayers with 2 qualifying children (percentage); 
Tax year 2002: AEITC: 53,721: (51%); 
Tax year 2002: EITC: 8,955,218: (43%); 
Tax year 2003: AEITC: 55,558: (52%); 
Tax year 2003: EITC: 9,139,643: (43%); 
Tax year 2004: AEITC: 52,448: (52%); 
Tax year 2004: EITC: 9,185,297: (42%). 

Total (percentage); 
Tax year 2002: AEITC: 104,733: (100%); 
Tax year 2002: EITC: 20,918,823: (100%); 
Tax year 2003: AEITC: 106,143; (100%); 
Tax year 2003: EITC: 21,431,377: (100%); 
Tax year 2004: AEITC: 100,613: (100%); 
Tax year 2004: EITC: 21,721,218: (100%). 

Source: GAO analysis of IRS data. 

Note: This table is based on the number of tax returns in the EITC 
database that had AEITC greater than $0 and zero, one, or two 
qualifying children. 

[End of table] 

Filing method of AEITC and EITC recipients: As noted in table 19, most 
tax returns that showed receiving an amount of AEITC were filed 
electronically, as were most tax returns which reported receipt of the 
EITC. For both AEITC and EITC, about 70 percent of recipients filed 
electronically for the 3 years we examined. 

Table 19: Filing Method of AEITC and EITC Recipients, Number and 
Percentage, Tax Years 2002 through 2004: 

Electronic; 
Tax year 2002: AEITC: 93,291: (68%); 
Tax year 2002: EITC: 14,030,564: (67%); 
Tax year 2003:AEITC: 100,085: (72%); 
Tax year 2003:EITC: 15,322,251: (71%); 
Tax year 2004: AEITC: 98,592: (78%); 
Tax year 2004: EITC: 16,269,437: (75%). 

Paper; 
Tax year 2002: AEITC: 44,477: (32%); 
Tax year 2002: EITC: 6,888,259: (33%); 
Tax year 2003:AEITC: 38,885: (28%); 
Tax year 2003:EITC: 6,109,126: (29%); 
Tax year 2004: AEITC: 28,228: (22%); 
Tax year 2004: EITC: 5,451,781: (25%). 

Total; 
Tax year 2002: AEITC: 137,768: (100%); 
Tax year 2002: EITC: 20,918,823: (100%); 
Tax year 2003:AEITC: 138,970: (100%); 
Tax year 2003:EITC: 21,431,377: (100%); 
Tax year 2004: AEITC: 126,820: (100%); 
Tax year 2004: EITC: 21,721,218: (100%). 

Source: GAO analysis of IRS data. 

Note: Tax years 2002 and 2003 include 5 and 4 duplicate Taxpayer 
Identification Numbers (TIN), respectively. The AEITC totals shown 
above do not equal the total number of individuals who filed a federal 
tax return for that year. This table only includes the number of tax 
returns that had an AEITC amount greater than $0. IRS did not identify 
the filing method used by individuals who filed a tax return but failed 
to report receipt of the AEITC. The total number of individuals who 
filed a federal tax return was 338,942, 316,959, and 307,251, for tax 
years 2002 through 2004 respectively. The EITC total represents the 
total number of taxpayers claiming EITC. 

[End of table] 

Geographic location of AEITC recipients in the valid subpopulation: Of 
the individuals in the valid subpopulation, use of the AEITC varied 
widely across the country. In all 3 tax years, Florida and Illinois had 
the most AEITC recipients (see fig. 7). 

Figure 7: Geographic Location of AEITC Recipients in the Valid 
Subpopulation, Tax Years 2002 through 2004: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

Note: AEITC geographic location is based upon information reported on 
the tax return. This figure includes the valid subpopulation. 

[A] These geographic locations had at least 1 but fewer than 10 
individuals who received the AEITC. 

[B] U.S insular areas include the three major U.S. territories, which 
are the U.S. Virgin Islands, American Samoa, Guam, and the two 
Commonwealths, which are the Commonwealth of Puerto Rico and the 
Commonwealth of Northern Mariana Islands. 

[C] Other includes Federated States of Micronesia, Marshall Islands, 
Palau, and Armed Forces mostly located outside the U.S. 

[D] Missing refers to instances where the geographic location was not 
available. 

[End of figure] 

Employer size, number of employers, number of forms W-2 employers 
issued to AEITC recipients and Total AEITC Dollars Reported on Forms W- 
2: Slightly more than 50,000 employers reported paying at least one 
employee AEITC in each tax years 2002 through 2004. Most of these 
employers were classified by IRS as small business/self employed and 
they issued more than half of all the Forms W-2 with AEITC (see table 
20). 

Table 20: Employer Size, Number of Employers, Number of Forms W-2 
Employers Issued to AEITC Recipients and Total Dollars Reported on 
Forms W-2, Tax Years 2002 through 2004: 

[See PDF for image]

Source: GAO analysis of IRS data. 

Note: Employers are defined by unique Employer Identification Numbers 
(EIN) and the classification of employer size is based on IRS criteria. 
Number of employers includes the valid, invalid name, and invalid 
number subpopulations. Number of Forms W-2 issued with AEITC includes 
all Forms W-2 issued that year reporting an amount of AEITC. 
Percentages may not add due to rounding. 

[End of table] 

[End of section] 

Appendix IV: Administrative and Legislative Changes to the Advance 
Earned Income Tax Credit Since 1990: 

IRS has made several administrative changes to the Advance Earned 
Income Tax Credit (AEITC) since the beginning of 1990 (see table 21). 
IRS described these changes in responses to recommendations in our 1992 
AEITC report and a 2003 Treasury Inspector General for Tax 
Administration report.[Footnote 61] 

Table 21: Administrative Changes to the AEITC since 1990: 

Year: 1990; 
Change: Added separate lines to the 1990 Forms 1040 and 1040A for 
reporting AEITC received. 

Year: 1991; 
Change: Revised Form W-4 to advise employees of the availability of the 
AEITC and to refer them to the Form W-5. 

Year: 1991-1992; 
Change: Revised Form W-2 and IRS Notice 797 to contain information on 
how to apply for the AEITC. 

Year: 1992; 
Change: Developed publicity materials and distributed them to the 
public; 
provided credit information in the Small Business Taxpayer Education 
Program; and increased outreach speaker seminar efforts, in cooperation 
with federal, state, and local officials and private organizations. 

Year: 1992; 
Change: Developed print and audiovisual products to inform the public 
about the AEITC. 

Year: 1992; 
Change: Expanded instructions on AEITC in "Circular E, Employer's Tax 
Guide.". 

Year: 1992; Change: Added examples to "Circular E, Employer's Tax 
Guide," explaining how to advance the EITC to employees and how to 
report amounts in advance. 

Year: 1992; 
Change: Clarified employer's responsibilities and liabilities in 
advancing the Earned Income Tax Credit (EITC) in the 1993 "Circular E, 
Employer's Tax Guide.". 

Year: 1993; 
Change: Expanded 1993 Form W-2 instructions concerning AEITC. 
Specifically, revised the Form W-2 to place a bold outline around the 
AEITC box of the employee's copy. 

Year: 2003; 
Change: Updated instructions that require tax examiners to compare the 
amount of AEITC shown on the tax return to the amount listed as AEITC 
on the Form W-2 and adjust the amounts when needed. 

Year: 2004; 
Change: Created an error condition in return processing when the amount 
of AEITC reported exceeds the maximum allowed for the year. 

Year: 2005; 
Change: Created an error condition in return processing for whenever 
AEITC is listed on the return without a corresponding entry for the 
EITC. 

Source: GAO analysis. 

Notes: Changes made between 1990 and 1993 were a result of GAO's 1992 
report and changes between 2003 and 2005 were a result of the Treasury 
Inspector General for Tax Administration's 2003 report. Some of the 
changes made between 1990 and 1993 occurred before our report was 
issued. 

[End of table] 

There have been two laws enacted since our 1992 report that include 
specific changes to the AEITC.[Footnote 62] First, the Omnibus Budget 
Reconciliation Act of 1993 (OBRA '93), did the following. 

1. Limited the amount of advance payment allowable in a taxable year to 
60 percent of the maximum credit available to a taxpayer with one 
qualifying child. 

2. Directed the IRS to notify taxpayers with qualifying children who 
receive a refund on account of the EITC that the credit may be 
available on an advance basis. The conference report accompanying OBRA 
'93 stated that after these notifications had been made for 2 taxable 
years, the Treasury Secretary was directed to study their effect on 
utilization of the advance payment mechanism and, based on the results 
of the study, the Secretary may recommend modifications to the 
notification program.[Footnote 63] 

Second, the U.S. Troop Readiness, Veteran's Care, Katrina Recovery, and 
Iraq Accountability Appropriations Act of 2007, which was enacted in 
late May 2007, calls for a study of AEITC use.[Footnote 64] The study 
is to be conducted by the Secretary of the Treasury for the Congress 
and is to include the benefits, costs, risks, and barriers to workers 
and to businesses (with a special emphasis on small businesses) if the 
AEITC included all recipients of the EITC (i.e., individuals without 
qualifying children). It also asks what steps would be necessary to 
implement such an inclusion. 

[End of section] 

Appendix V: Additional Analyses of Advance Earned Income Tax Credit 
Noncompliance: 

We identified additional areas of noncompliance for AEITC recipients 
during tax years 2002 through 2004. We also examined demographic 
characteristics of our noncompliant subpopulations, including the 
invalid name, invalid number, and dollar limit subpopulations. Our 
analyses revealed the following: (1) some consecutive AEITC recipients 
had an invalid SSN, (2) most consecutive AEITC recipients filed a tax 
return, but did not report the correct AEITC amount on the tax return, 
(3) AEITC recipients with an invalid SSN received little money, (4) 
most AEITC recipients with an invalid SSN received one to two Forms W- 
2, (5) AEITC recipients in the invalid subpopulations lived in various 
geographic locations, and (6) AEITC was paid in excess of yearly 
maximum limits. As with the data in Appendix III, further analyses of 
this data may provide information on noncompliance characteristics 
potentially useful for IRS enforcement efforts. 

Each of the tables provides additional information about AEITC 
recipients not previously discussed. Where possible, we compared AEITC 
recipients to EITC recipients. The characteristics are organized by a 
declining AEITC population size, similar to the organization in 
appendix III. Most analyses include the invalid name and invalid number 
subpopulations. The subpopulations are described in the scope and 
methodology section (see app. II). 

Some consecutive AEITC recipients had an invalid SSN: About 98,000 
individuals received the AEITC consecutively in each of the 3 years, 
tax years 2002 through 2004, and received an average of about $56 
million. About a quarter of these individuals had an invalid SSN and 
received an average of approximately $16 million in AEITC. 
Additionally, nearly 15 percent of consecutive users had an invalid SSN 
and did not file a federal tax return, receiving an average of about $9 
million in AEITC (see table 22). 

Table 22: Number of Individuals and AEITC Dollars Received by AEITC 
Recipients and Individuals Who Elected the AEITC Consecutively Using an 
Invalid SSN and an Invalid SSN and Not Filing, Tax Years 2002 through 
2004: 

AEITC recipients; 
Tax year 2002: Individuals: 531,799; 
Tax year 2002: Dollars: $144,547,006; 
Tax Year 2003: Individuals: 503,382; 
Tax Year 2003: Dollars: $143,205,901; 
Tax Year 2004: Individuals: 507,957; 
Tax Year 2004: Dollars: $150,084,267. 

AEITC consecutive recipients[A]; 
Tax year 2002: Individuals: 97,998: (18%); 
Tax year 2002: Dollars: $53,272,152: (37%); 
Tax Year 2003: Individuals: 97,998: (19%); 
Tax Year 2003: Dollars: $60,849,730: (42%); 
Tax Year 2004: Individuals: 97,998: (19%); 
Tax Year 2004: Dollars: $52,698,947: (35%). 

AEITC consecutive recipients with invalid SSN (percentage)[B]; 
Tax year 2002: Individuals: 23,175: (24%); 
Tax year 2002: Dollars: $14,837,041: (28%); 
Tax Year 2003: Individuals: 23,175; (24%); 
Tax Year 2003: Dollars: $17,217,171: (28%); 
Tax Year 2004: Individuals: 23,175: (24%); 
Tax Year 2004: Dollars: $14,798,025: (28%). 

AEITC consecutive recipients with invalid SSN and did not file federal 
tax return (percentage)[B]; 
Tax year 2002: Individuals: 13,181; (13%); 
Tax year 2002: Dollars: $8,593,857: (16%); 
Tax Year 2003: Individuals: 13,153: (13%); 
Tax Year 2003: Dollars: $9,972,631: (16%); 
Tax Year 2004: Individuals: 13,348: (14%); 
Tax Year 2004: Dollars: $8,722,592: (17%). 

Source: GAO analysis of IRS data. 

Note: The number of AEITC consecutive recipients represents the total 
number of individuals who received the AEITC over each of the 3 years. 
The associated dollars represent the amount those individuals received 
each year. 

[A] Percentage is out of AEITC recipients and dollars. This analysis 
includes the valid, invalid name, and invalid number subpopulations. 

[B] Percentage is out of AEITC consecutive recipients and dollars. This 
analysis includes the invalid name and invalid number subpopulations. 

Most consecutive AEITC recipients filed a tax return, but did not 
report the correct AEITC amount on the tax return: About 98,000 
individuals received the AEITC consecutively in each of the 3 tax 
years, 2002 through 2004. More than half of these individuals filed a 
tax return. Of those who filed, about half reported the same AEITC 
amount on the tax return as shown on the Form W-2 (i.e., matched). Of 
the mismatches, the majority did not report receipt of the AEITC (see 
table 23). 

[End of table] 

Table 23: Number of AEITC Consecutive Recipients with Matches and 
Mismatches between Forms W-2 and Filed Tax Returns, Tax Years 2002 
through 2004: 

[See PDF for image] 

[End of table] 

Source: GAO analysis of IRS data. 

Note: Percentages may not add due to rounding. This table includes 
individuals who filed a tax return in the valid and invalid name 
subpopulations. 

[A] We considered a match to be anything plus or minus a dollar on the 
Form W-2 in order to allow for taxpayer rounding. We tested the 
sensitivity of this result by using a difference between the AEITC 
reported on the Form W-2 and the tax return of up to $100 and the 
results were similar. Percentage is of number of consecutive AEITC 
recipients who filed federal tax returns. 

[B] Percentage is of number of mismatches due to nonreporting of AEITC 
for consecutive recipients and number of consecutive AEITC recipients 
who filed federal tax returns. 

AEITC recipients with an invalid Social Security number (SSN) received 
little money: Most AEITC recipients who had a Form(s) W-2 with an 
invalid SSN obtained $100 or less of AEITC (see table 24). These data 
are consistent with the overall AEITC population, as previously noted, 
where about half of all recipients received less than $100 and 80 
percent received $500 or less for the year. 

[End of figure] 

Table 24: Amount of AEITC Received Per Invalid Form W-2, Number and 
Percentage, Tax Years 2002 through 2004: 

[See PDF for image]

Source: GAO analysis of IRS data. 

Note: This table includes the invalid number subpopulation. 

[A] Above the maximum is possible because this table presents 
aggregated Forms W-2 and if an individual received more than one Form W-
2, the total received could be above the yearly maximum. 

[End of table] 

Most AEITC recipients with an invalid SSN received 1 to 2 Forms W-2: 
Most recipients who had an invalid SSN received 1 to 2 Forms W-2 
reporting AEITC. For example, in tax year 2002, 6,223 individuals 
received two Forms W-2 that reported AEITC. These resulted in a total 
of 12,466 Forms W-2 equaling $3,261,327 in AEITC, with an average of 
$262 in AEITC per Form W-2 (see tables 25, 26, and 27). 

Table 25: Number of Individuals Receiving a Form(s) W-2, Total Number 
of Form(s) W-2 Reporting AEITC, Total Associated and Average Dollars 
Received Relative to Number of Form(s) W-2 Reporting AEITC for AEITC 
Recipients with an Invalid SSN on the Form(s) W-2, Tax Year 2002: 

A: Number of Forms W-2 an individual received reporting AEITC: 1; 
B: Number of individuals receiving the number of Forms W-2: 22,812; 
C: Total number of Forms W-2 reporting AEITC[A]: 22,812; 
D: Total associated dollars on Forms W-2: $5,588,671; 
E: Average dollars per Form W-2[B]: $245. 

A: Number of Forms W-2 an individual received reporting AEITC: 2; 
B: Number of individuals receiving the number of Forms W-2: 6,233; 
C: Total number of Forms W-2 reporting AEITC[A]: 12,466; 
D: Total associated dollars on Forms W-2: $3,261,327; 
E: Average dollars per Form W-2[B]: $262. 

A: Number of Forms W-2 an individual received reporting AEITC: 3; 
B: Number of individuals receiving the number of Forms W-2: 1,954; 
C: Total number of Forms W-2 reporting AEITC[A]: 5,862; 
D: Total associated dollars on Forms W-2: $1,290,955; 
E: Average dollars per Form W-2[B]: $220. 

A: Number of Forms W-2 an individual received reporting AEITC: 4; 
B: Number of individuals receiving the number of Forms W-2: 874; 
C: Total number of Forms W-2 reporting AEITC[A]: 3,496; 
D: Total associated dollars on Forms W-2: $689,595; 
E: Average dollars per Form W-2[B]: $197. 

A: Number of Forms W-2 an individual received reporting AEITC: 5; 
B: Number of individuals receiving the number of Forms W-2: 332; 
C: Total number of Forms W-2 reporting AEITC[A]: 1,660; 
D: Total associated dollars on Forms W-2: $287,912; 
E: Average dollars per Form W-2[B]: $173. 

A: Number of Forms W-2 an individual received reporting AEITC: 6; 
B: Number of individuals receiving the number of Forms W-2: 142; 
C: Total number of Forms W-2 reporting AEITC[A]: 852; 
D: Total associated dollars on Forms W-2: $138,384; 
E: Average dollars per Form W-2[B]: $162. 

A: Number of Forms W-2 an individual received reporting AEITC: 7; 
B: Number of individuals receiving the number of Forms W-2: 74; 
C: Total number of Forms W-2 reporting AEITC[A]: 518; 
D: Total associated dollars on Forms W-2: $84,688; 
E: Average dollars per Form W-2[B]: $163. 

A: Number of Forms W-2 an individual received reporting AEITC: 8; 
B: Number of individuals receiving the number of Forms W-2: 37; 
C: Total number of Forms W-2 reporting AEITC[A]: 296; 
D: Total associated dollars on Forms W-2: $46,962; 
E: Average dollars per Form W-2[B]: $159. 

A: Number of Forms W-2 an individual received reporting AEITC: 9; 
B: Number of individuals receiving the number of Forms W-2: 10; 
C: Total number of Forms W-2 reporting AEITC[A]: 90; 
D: Total associated dollars on Forms W-2: $15,429; 
E: Average dollars per Form W-2[B]: $171. 

A: Number of Forms W-2 an individual received reporting AEITC: 10; 
B: Number of individuals receiving the number of Forms W-2: 7; 
C: Total number of Forms W-2 reporting AEITC[A]: 70; 
D: Total associated dollars on Forms W-2: $9,601; 
E: Average dollars per Form W-2[B]: $137. 

A: Number of Forms W-2 an individual received reporting AEITC: 11; 
B: Number of individuals receiving the number of Forms W-2: 7; 
C: Total number of Forms W-2 reporting AEITC[A]: 77; 
D: Total associated dollars on Forms W-2: $12,577; 
E: Average dollars per Form W-2[B]: $163. 

A: Number of Forms W-2 an individual received reporting AEITC: 12; 
B: Number of individuals receiving the number of Forms W-2: 3; 
C: Total number of Forms W-2 reporting AEITC[A]: 36; 
D: Total associated dollars on Forms W-2: $6,073; 
E: Average dollars per Form W-2[B]: $169. 

A: Number of Forms W-2 an individual received reporting AEITC: 13; 
B: Number of individuals receiving the number of Forms W-2: 4; 
C: Total number of Forms W-2 reporting AEITC[A]: 52; 
D: Total associated dollars on Forms W-2: $11,805; 
E: Average dollars per Form W-2[B]: $227. 

A: Number of Forms W-2 an individual received reporting AEITC: 14; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 14; 
D: Total associated dollars on Forms W-2: $6,246; 
E: Average dollars per Form W-2[B]: $446. 

A: Number of Forms W-2 an individual received reporting AEITC: 15; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 15; 
D: Total associated dollars on Forms W-2: $5,156; 
E: Average dollars per Form W-2[B]: $344. 

A: Number of Forms W-2 an individual received reporting AEITC: 16; 
B: Number of individuals receiving the number of Forms W-2: 2; 
C: Total number of Forms W-2 reporting AEITC[A]: 32; 
D: Total associated dollars on Forms W-2: $8,271; 
E: Average dollars per Form W-2[B]: $258. 

A: Number of Forms W-2 an individual received reporting AEITC: 17; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 17; 
D: Total associated dollars on Forms W-2: $3,893; 
E: Average dollars per Form W-2[B]: $229. 

A: Number of Forms W-2 an individual received reporting AEITC: 19; 
B: Number of individuals receiving the number of Forms W-2: 2; 
C: Total number of Forms W-2 reporting AEITC[A]: 38; 
D: Total associated dollars on Forms W-2: $10,695; 
E: Average dollars per Form W-2[B]: $281. 

A: Number of Forms W-2 an individual received reporting AEITC: 20; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 20; 
D: Total associated dollars on Forms W-2: $6,978; 
E: Average dollars per Form W-2[B]: $349. 

A: Number of Forms W-2 an individual received reporting AEITC: 23; 
B: Number of individuals receiving the number of Forms W-2: 2; 
C: Total number of Forms W-2 reporting AEITC[A]: 46; 
D: Total associated dollars on Forms W-2: $8,101; 
E: Average dollars per Form W-2[B]: $176. 

A: Number of Forms W-2 an individual received reporting AEITC: 28; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 28; 
D: Total associated dollars on Forms W-2: $6,074; 
E: Average dollars per Form W-2[B]: $217. 

A: Number of Forms W-2 an individual received reporting AEITC: 935; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 935; 
D: Total associated dollars on Forms W-2: $83,211; 
E: Average dollars per Form W-2[B]: $89. 

A: Number of Forms W-2 an individual received reporting AEITC: Total 
Forms W-2 with invalid SSNs; 
B: Number of individuals receiving the number of Forms W-2: 32,501; 
C: Total number of Forms W-2 reporting AEITC[A]: 49,432; 
D: Total associated dollars on Forms W-2: $11,582,604; 
E: Average dollars per Form W-2[B]: $234. 

Source: GAO analysis of IRS data. 

Note: This table includes the invalid number subpopulation. 

[A] This number is derived by multiplying column A by column B. 

[B] This amount is derived by dividing column D by column C. 

[End of table] 

Table 26: Number of Individuals Receiving a Form(s) W-2, Total Number 
of Form(s) W-2 Reporting AEITC, Total Associated and Average Dollars 
Received, Relative to Number of Form(s) W-2 Reporting AEITC for AEITC 
Recipients with an Invalid SSN on the Form(s) W-2, Tax Year 2003: 

A: Number of Forms W-2 an individual received reporting AEITC: 1; 
B: Number of individuals receiving the number of Forms W-2: 22,474; 
C: Total number of Forms W-2 reporting AEITC[A]: 22,474; 
D: Total associated dollars on Forms W-2: $6,787,458; 
E: Average dollars per Forms W-2[B]: $302. 

A: Number of Forms W-2 an individual received reporting AEITC: 2; 
B: Number of individuals receiving the number of Forms W-2: 5,424; 
C: Total number of Forms W-2 reporting AEITC[A]: 10,848; 
D: Total associated dollars on Forms W-2: $2,689,186; 
E: Average dollars per Forms W-2[B]: $248. 

A: Number of Forms W-2 an individual received reporting AEITC: 3; 
B: Number of individuals receiving the number of Forms W-2: 1,660; 
C: Total number of Forms W-2 reporting AEITC[A]: 4,980; 
D: Total associated dollars on Forms W-2: $1,076,219; 
E: Average dollars per Forms W-2[B]: $216. 

A: Number of Forms W-2 an individual received reporting AEITC: 4; 
B: Number of individuals receiving the number of Forms W-2: 600; 
C: Total number of Forms W-2 reporting AEITC[A]: 2,400; 
D: Total associated dollars on Forms W-2: $478,420; 
E: Average dollars per Forms W-2[B]: $199. 

A: Number of Forms W-2 an individual received reporting AEITC: 5; 
B: Number of individuals receiving the number of Forms W-2: 259; 
C: Total number of Forms W-2 reporting AEITC[A]: 1,295; 
D: Total associated dollars on Forms W-2: $222,288; 
E: Average dollars per Forms W-2[B]: $172. 

A: Number of Forms W-2 an individual received reporting AEITC: 6; 
B: Number of individuals receiving the number of Forms W-2: 94; 
C: Total number of Forms W-2 reporting AEITC[A]: 564; 
D: Total associated dollars on Forms W-2: $107,704; 
E: Average dollars per Forms W-2[B]: $191. 

A: Number of Forms W-2 an individual received reporting AEITC: 7; 
B: Number of individuals receiving the number of Forms W-2: 41; 
C: Total number of Forms W-2 reporting AEITC[A]: 287; 
D: Total associated dollars on Forms W-2: $46,869; 
E: Average dollars per Forms W-2[B]: $163. 

A: Number of Forms W-2 an individual received reporting AEITC: 8; 
B: Number of individuals receiving the number of Forms W-2: 22; 
C: Total number of Forms W-2 reporting AEITC[A]: 176; 
D: Total associated dollars on Forms W-2: $32,510; 
E: Average dollars per Forms W-2[B]: $185. 

A: Number of Forms W-2 an individual received reporting AEITC: 9; 
B: Number of individuals receiving the number of Forms W-2: 17; 
C: Total number of Forms W-2 reporting AEITC[A]: 153; 
D: Total associated dollars on Forms W-2: $30,123; 
E: Average dollars per Forms W-2[B]: $197. 

A: Number of Forms W-2 an individual received reporting AEITC: 10; 
B: Number of individuals receiving the number of Forms W-2: 6; 
C: Total number of Forms W-2 reporting AEITC[A]: 60; 
D: Total associated dollars on Forms W-2: $14,293; 
E: Average dollars per Forms W-2[B]: $238. 

A: Number of Forms W-2 an individual received reporting AEITC: 11; 
B: Number of individuals receiving the number of Forms W-2: 4; 
C: Total number of Forms W-2 reporting AEITC[A]: 44; 
D: Total associated dollars on Forms W-2: $9,769; 
E: Average dollars per Forms W-2[B]: $222. 

A: Number of Forms W-2 an individual received reporting AEITC: 12; 
B: Number of individuals receiving the number of Forms W-2: 5; 
C: Total number of Forms W-2 reporting AEITC[A]: 60; 
D: Total associated dollars on Forms W-2: $14,794; 
E: Average dollars per Forms W-2[B]: $247. 

A: Number of Forms W-2 an individual received reporting AEITC: 13; 
B: Number of individuals receiving the number of Forms W-2: 5; 
C: Total number of Forms W-2 reporting AEITC[A]: 65; 
D: Total associated dollars on Forms W-2: $20,860; 
E: Average dollars per Forms W-2[B]: $321. 

A: Number of Forms W-2 an individual received reporting AEITC: 14; 
B: Number of individuals receiving the number of Forms W-2: 3; 
C: Total number of Forms W-2 reporting AEITC[A]: 42; 
D: Total associated dollars on Forms W-2: $10,857; 
E: Average dollars per Forms W-2[B]: $259. 

A: Number of Forms W-2 an individual received reporting AEITC: 15; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 15; 
D: Total associated dollars on Forms W-2: $1,798; 
E: Average dollars per Forms W-2[B]: $120. 

A: Number of Forms W-2 an individual received reporting AEITC: 16; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 16; 
D: Total associated dollars on Forms W-2: $4,378; 
E: Average dollars per Forms W-2[B]: $274. 

A: Number of Forms W-2 an individual received reporting AEITC: 17; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 17; 
D: Total associated dollars on Forms W-2: $10,284; 
E: Average dollars per Forms W-2[B]: $605. 

A: Number of Forms W-2 an individual received reporting AEITC: 19; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Forms W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 20; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 20; 
D: Total associated dollars on Forms W-2: $4,468; 
E: Average dollars per Forms W-2[B]: $223. 

A: Number of Forms W-2 an individual received reporting AEITC: 22; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Forms W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 23; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Forms W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 26; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 26; 
D: Total associated dollars on Forms W-2: $8,584; 
E: Average dollars per Forms W-2[B]: $330. 

A: Number of Forms W-2 an individual received reporting AEITC: 28; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 28; 
D: Total associated dollars on Forms W-2: $9,513; 
E: Average dollars per Forms W-2[B]: $340. 

A: Number of Forms W-2 an individual received reporting AEITC: 31; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 31; 
D: Total associated dollars on Forms W-2: $9,404; 
E: Average dollars per Forms W-2[B]: $303. 

A: Number of Forms W-2 an individual received reporting AEITC: 38; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Forms W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 643; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 643; 
D: Total associated dollars on Forms W-2: $60,948; 
E: Average dollars per Forms W-2[B]: $95. 

A: Number of Forms W-2 an individual received reporting AEITC: 935; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Forms W-2[B]: 0. 

A: Number of Forms W-2 an individual received reporting AEITC: 1,088; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Forms W-2[B]: 0. 

A: Number of Forms W-2 an individual received reporting AEITC: Total; 
B: Number of individuals receiving the number of Forms W-2: 30,622; 
C: Total number of Forms W-2 reporting AEITC[A]: 44,244; 
D: Total associated dollars on Forms W-2: 11,650,727; 
E: Average dollars per Forms W-2[B]: $263. 

Source: GAO analysis of IRS data. 

Note: This table includes the invalid number subpopulation. 

[A] This number is derived by multiplying column A by column B. 

[B] This amount is derived by dividing column D by column C. 

[End of table] 

Table 27: Number of Individuals Receiving a Form(s) W-2, Total Number 
of Form(s) W-2 Reporting AEITC, Total Associated and Average Dollars 
Received, Relative to Number of Form(s) W-2 Reporting AEITC for AEITC 
Recipients with an Invalid SSN on the Form(s) W-2, Tax Year 2004: 

A: Number of Forms W-2 an individual received reporting AEITC: 1; 
B: Number of individuals receiving the number of Forms W-2: 24,853; 
C: Total number of Forms W-2 reporting AEITC[A]: 24,853; 
D: Total associated dollars on Forms W-2: $7,602,238; 
E: Average dollars per Form W-2[B]: $306. 

A: Number of Forms W-2 an individual received reporting AEITC: 2; 
B: Number of individuals receiving the number of Forms W-2: 4,776; 
C: Total number of Forms W-2 reporting AEITC[A]: 9,552; 
D: Total associated dollars on Forms W-2: $2,663,173; 
E: Average dollars per Form W-2[B]: $279. 

A: Number of Forms W-2 an individual received reporting AEITC: 3; 
B: Number of individuals receiving the number of Forms W-2: 1,452; 
C: Total number of Forms W-2 reporting AEITC[A]: 4,356; 
D: Total associated dollars on Forms W-2: $1,022,262; 
E: Average dollars per Form W-2[B]: $235. 

A: Number of Forms W-2 an individual received reporting AEITC: 4; 
B: Number of individuals receiving the number of Forms W-2: 483; 
C: Total number of Forms W-2 reporting AEITC[A]: 1,932; 
D: Total associated dollars on Forms W-2: $385,252; 
E: Average dollars per Form W-2[B]: $199. 

A: Number of Forms W-2 an individual received reporting AEITC: 5; 
B: Number of individuals receiving the number of Forms W-2: 187; 
C: Total number of Forms W-2 reporting AEITC[A]: 935; 
D: Total associated dollars on Forms W-2: $178,214; 
E: Average dollars per Form W-2[B]: $191. 

A: Number of Forms W-2 an individual received reporting AEITC: 6; 
B: Number of individuals receiving the number of Forms W-2: 57; 
C: Total number of Forms W-2 reporting AEITC[A]: 342; 
D: Total associated dollars on Forms W-2: $72,679; 
E: Average dollars per Form W-2[B]: $213. 

A: Number of Forms W-2 an individual received reporting AEITC: 7; 
B: Number of individuals receiving the number of Forms W-2: 27; 
C: Total number of Forms W-2 reporting AEITC[A]: 189; 
D: Total associated dollars on Forms W-2: $47,036; 
E: Average dollars per Form W-2[B]: $249. 

A: Number of Forms W-2 an individual received reporting AEITC: 8; 
B: Number of individuals receiving the number of Forms W-2: 13; 
C: Total number of Forms W-2 reporting AEITC[A]: 104; 
D: Total associated dollars on Forms W-2: $30,029; 
E: Average dollars per Form W-2[B]: $289. 

A: Number of Forms W-2 an individual received reporting AEITC: 9; 
B: Number of individuals receiving the number of Forms W-2: 8; 
C: Total number of Forms W-2 reporting AEITC[A]: 72; 
D: Total associated dollars on Forms W-2: $15,628; 
E: Average dollars per Form W-2[B]: $217. 

A: Number of Forms W-2 an individual received reporting AEITC: 10; 
B: Number of individuals receiving the number of Forms W-2: 4; 
C: Total number of Forms W-2 reporting AEITC[A]: 40; 
D: Total associated dollars on Forms W-2: $9,060; 
E: Average dollars per Form W-2[B]: $227. 

A: Number of Forms W-2 an individual received reporting AEITC: 11; 
B: Number of individuals receiving the number of Forms W-2: 5; 
C: Total number of Forms W-2 reporting AEITC[A]: 55; 
D: Total associated dollars on Forms W-2: $9,027; 
E: Average dollars per Form W-2[B]: $164. 

A: Number of Forms W-2 an individual received reporting AEITC: 12; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 12; 
D: Total associated dollars on Forms W-2: $3,178; 
E: Average dollars per Form W-2[B]: $265. 

A: Number of Forms W-2 an individual received reporting AEITC: 13; 
B: Number of individuals receiving the number of Forms W-2: 2; 
C: Total number of Forms W-2 reporting AEITC[A]: 26; 
D: Total associated dollars on Forms W-2: $9,049; 
E: Average dollars per Form W-2[B]: $348. 

A: Number of Forms W-2 an individual received reporting AEITC: 14; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Form W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 15; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 15; 
D: Total associated dollars on Forms W-2: $5,077; 
E: Average dollars per Form W-2[B]: $338. 

A: Number of Forms W-2 an individual received reporting AEITC: 16; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Form W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 17; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Form W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 19; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 19; 
D: Total associated dollars on Forms W-2: $5,604; 
E: Average dollars per Form W-2[B]: $295. 

A: Number of Forms W-2 an individual received reporting AEITC: 20; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 20; 
D: Total associated dollars on Forms W-2: $9,567; 
E: Average dollars per Form W-2[B]: $478. 

A: Number of Forms W-2 an individual received reporting AEITC: 22; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 22; 
D: Total associated dollars on Forms W-2: $5,076; 
E: Average dollars per Form W-2[B]: $231. 

A: Number of Forms W-2 an individual received reporting AEITC: 23; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Form W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 26; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 26; 
D: Total associated dollars on Forms W-2: $6,837; 
E: Average dollars per Form W-2[B]: $263. 

A: Number of Forms W-2 an individual received reporting AEITC: 28; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Form W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 31; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Form W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 38; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 38; 
D: Total associated dollars on Forms W-2: $9,575; 
E: Average dollars per Form W-2[B]: $252. 

A: Number of Forms W-2 an individual received reporting AEITC: 643; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Form W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 935; 
B: Number of individuals receiving the number of Forms W-2: 0; 
C: Total number of Forms W-2 reporting AEITC[A]: 0; 
D: Total associated dollars on Forms W-2: $0; 
E: Average dollars per Form W-2[B]: $0. 

A: Number of Forms W-2 an individual received reporting AEITC: 1,088; 
B: Number of individuals receiving the number of Forms W-2: 1; 
C: Total number of Forms W-2 reporting AEITC[A]: 1,088; 
D: Total associated dollars on Forms W-2: $78,141; 
E: Average dollars per Form W-2[B]: $72. 

A: Number of Forms W-2 an individual received reporting AEITC: Total; 
B: Number of individuals receiving the number of Forms W-2: 31,875; 
C: Total number of Forms W-2 reporting AEITC[A]: 43,696; 
D: Total associated dollars on Forms W-2: $12,166,702; 
E: Average dollars per Form W-2[B]: $278. 

Source: GAO analysis of IRS data. 

Note: This table includes the invalid number subpopulation. 

[A] This number is derived by multiplying column A by column B. 

[B] This amount is derived by dividing column D by column C. 

[End of table] 

AEITC recipients in the invalid subpopulations lived in various 
geographic locations: Use of the AEITC varied widely across the country 
for individuals in the invalid number and invalid name subpopulations. 
For the invalid name subpopulation, in all 3 tax years, California and 
Illinois had the most Forms W-2 reporting AEITC and for the invalid 
number subpopulation, in all 3 tax years, Florida and Illinois had the 
most (see figs. 8 and 9). 

Figure 8: Geographic Location of AEITC Recipients in the Invalid Name 
Subpopulation, Tax Years 2002 through 2004: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

Note: AEITC geographic location is based upon information reported on 
the tax return. This figure includes the invalid name subpopulation. 

[A] These geographic locations had at least one but fewer than 10 
individuals who received the AEITC. 

[B] U.S insular areas include the three major U.S. territories, which 
are the U.S. Virgin Islands, American Samoa, Guam, and the two 
Commonwealths, which are the Commonwealth of Puerto Rico and the 
Commonwealth of Northern Mariana Islands. 

[C] Other includes Federated States of Micronesia, Marshall Islands, 
Palau, and Armed Forces mostly located outside the U.S. 

[D] Missing refers to instances where the geographic location was not 
available. 

[End of figure] 

Figure 9: Geographic Location of AEITC Recipients in the Invalid Number 
Subpopulation, Tax Years 2002 through 2004: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

Note: AEITC geographic location is based upon information reported on 
the tax return. This figure includes the invalid number subpopulation. 

[A] These geographic locations had at least 1 but fewer than 10 
individuals who received the AEITC. 

[B] U.S insular areas includes the three major U.S. territories, which 
are the U.S. Virgin Islands, American Samoa, Guam, and the two 
Commonwealths, which are the Commonwealth of Puerto Rico and the 
Commonwealth of Northern Mariana Islands. 

[C] Missing refers to instances where the geographic location was not 
available. 

[End of figure] 

AEITC was paid in excess of yearly maximum limits: A total of almost 
12,000 Forms W-2, reporting about $64 million, showed AEITC paid above 
the yearly maximum between tax years 2002 through 2004 (see figs. 10 
and 11). Specifically, in tax year 2002 there were 6,408 Forms W-2 
above the yearly maximum reporting almost $44 million; 
2,690 in tax year 2003, reporting over $7 million; 
and 2,768 in tax year 2004, reporting almost $13 million. As noted in 
figure 10, most Forms W-2 above the yearly maximum were between $1 
above the limit and $5,000. An individual receiving the AEITC was 
eligible to obtain a maximum yearly amount of $1,503 in tax year 2002, 
$1,528 in tax year 2003, and $1,563 in tax year 2004. 

Figure 10: Number of Forms W-2 Received with Amounts above the Yearly 
AEITC Maximum, Tax Years 2002 through 2004: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

Note: This excludes 2 Forms W-2 in 2003 that when totaled, reported 
about $1 billion paid in AEITC. This figure includes the dollar limit 
subpopulation. 

[End of figure] 

Figure 11: AEITC Dollars Reported on Form W-2s above the Yearly AEITC 
Maximum, Tax Years 2002-2004: 

[See PDF for image] 

Source: GAO analysis of IRS data. 

Note: This excludes 2 Forms W-2 in 2003 that when totaled, reported 
about $1 billion paid in AEITC. This figure includes the dollar limit 
subpopulation. 

[End of figure] 

[End of section] 

Appendix VI: Comments from the Internal Revenue Service: 

Commissioner:

Department Of The Treasury: 
Internal Revenue Service: 
Washington, D.C. 20224: 

July 18, 2007

Mr. Michael Brostek: 
Director, Tax Issues Strategic Issues Team: 
U.S. Government Accountability Office: 
441 G Street, N.W.: 
Washington, D.C. 20548: 

Dear Mr. Brostek: 

I am writing in response to the draft Government Accountability Office 
(GAO) audit report entitled Advance Earned Income Tax Credit, Low Use 
and Dollars Impede Efforts to Reduce Noncompliance (GAO-07-1110). I 
reviewed the report and generally agree with its findings and 
recommendations. The report accurately reflects the difficulty the IRS 
faces to increase participation in the Advance Earned Income Tax Credit 
(AEITC) payment option among eligible taxpayers. Your report also 
appropriately acknowledges the challenges involved in efforts to reduce 
noncompliance among the small percentage that receive AEITC. 

The IRS realizes the importance and potential benefits of AEITC as an 
immediate reward for work effort and as a help in meeting the day-to-
day needs of low income workers. However, both your analysis and our 
own experience suggest that most taxpayers prefer to receive the earned 
income tax credit (EITC) as a lump sum rather than in smaller amounts 
meted out over time. As a result, despite our efforts to promote AEITC 
and to educate employers and taxpayers on the benefits of this option 
participation remains low. Currently, 3 percent of the potentially 
eligible population opt for AEITC, and this percentage has grown only 
slightly since your 1992 report on the subject (EIC Advance Payment 
(GAOIGGD 92-26)). 

Your report also cites the high degree of noncompliance among the small 
number of participants receiving AEITC. However, because the 
participation rate for this program is low and half of all advance 
payment workers receive $100 or less in AEITC during the year, 
noncompliance in this area represents a very small fraction of the 
overall tax gap. For these reasons, the IRS must carefully weigh its 
efforts to address AEITC error against opportunities to address other 
areas of noncompliance. 

I appreciate the continued and valuable support from you and your staff 
on this issue. Detailed comments on your specific recommendation are 
enclosed. If you have any questions or would like to discuss this 
response in more detail, please contact David R. Williams, Director, 
Electronic Tax Administration and Refundable Credits, at (202) 622-
7990. 

Sincerely, 

Signed by: 
Kevin M. Brown: 
Acting Commissioner: 

Enclosure: 

Recommendation: 

The Commissioner of Internal Revenue should analyze whether any of the 
following options could cost effectively and significantly reduce AEITC 
noncompliance: 

* sending potentially noncompliant AEITC recipients soft notices, such 
as to nonfilers whose Forms W-2 show that they received AEITC and 
filers who misreported the amount they received and/or whose SSN and 
name do not match;
* requiring employers to verify the SSN of employees seeking AEITC; or  
* requiring employers to submit Form(s) W-5 to IRS and IRS creating and 
maintaining a database for these forms. 

To better identify the costs and implementation issues as well as the 
likelihood for these or other options to reduce AEITC noncompliance, 
where practical, the Commissioner should test these options to make a 
more fully informed judgment about whether any would be worthwhile. 

If the Commissioner determines that none of these options would be cost 
effective and no other remedies are viable, then the Treasury Secretary 
should inform the Congress of this and provide Treasury's opinion about 
whether the AEITC should be retained. 

Response: 

IRS agrees to conduct further cost/benefits analyses of the three 
proposed options. We will consider each within the context of the 
number of taxpayers affected, dollars involved (average per case and in 
aggregate), and the trade-offs from the return on investment realized 
by addressing other areas of the tax gap. 

Of the three compliance treatments recommended, we believe the soft 
notice option is the most viable, despite the known limitations cited 
in your report. The high AEITC turnover and nonfiler rates may make 
this option less effective and more costly than other successful IRS 
soft notice tests. While we agree the soft notice option may produce 
first year compliance improvements, it is unclear that it will 
significantly reduce overall AEITC noncompliance or substantially 
increase revenue protected. Important parts of our cost-benefit 
analysis of this option must also include collection potential, 
taxpayer burden, and costs associated with provisions for free or low 
cost alternatives for return preparation assistance. 

The option to require employers to verify SSNs of employees seeking 
AEITC will require legislation and new inter-agency coordination of 
roles and responsibilities between IRS and the Social Security 
Administration (SSA). Your report appropriately cites other concerns 
raised by IRS and SSA, such as data accuracy, additional employer 
responsibilities and burden, adverse effects on participation, 
potential for inconsistent treatment of workers due to the disparity 
between antidiscrimination laws and legal requirements for verifying 
employment eligibility status, and the need for additional agency 
resources. The option to require employers to submit Forms W-5 to IRS 
and for the IRS to create and maintain a database for these forms will 
generate new employer compliance issues, as well as concerns similar to 
those for the SSN verification option. Although we will conduct further 
analyses of these two options, we believe these challenges may preclude 
any meaningful near-term testing or implementation. 

Finally, our cost/benefits analyses of your recommended options will be 
conducted in tandem with a study of AEITC use required by the U.S. 
Troop Readiness, Veteran's Care, Katrina Recovery, and Iraq 
Accountability Appropriations Act of 2007 (PL 110-28) enacted in late 
May 2007. The study will include the benefits, costs, risks, and 
barriers to workers and to businesses (with a special emphasis on small 
businesses) if the AEITC included all recipients of the EITC (i.e., 
individuals without qualifying children), and the steps necessary for 
implementation. This report will be reviewed and approved by the 
Treasury Department and is due to the Congress in November 2007.

[End of section] 

Appendix VII: GAO Contacts and Staff Acknowledgments: 

GAO Contacts: 

Michael Brostek, (202) 512-9110 Libby Mixon, (404) 679-1900 Joanna 
Stamatiades, (404) 679-1900: 

Acknowledgments: 

In addition to those named above, Blake Ainsworth, Frances Cook, James 
Cook, Rebecca Gambler, Evan Gilman, George Guttman, Donna Miller, 
Cheryl Peterson, Michael Rose, Steve Sebastian, Daniel Schwimer, 
Richard Stana, James Ungvarsky, Michael Volpe, and Paul Wright made key 
contributions to this report. 

Footnotes: 

[1] The other credit with an advance option is the Health Coverage Tax 
Credit, which pays 65 percent of qualified health plan premiums for 
eligible trade-affected workers and certain Pension Benefit Guaranty 
Corporation benefit recipients. 

[2] In tax year 2004, about 22 million individuals received about $40 
billion in EITC payments from the IRS. Any portion of the EITC that an 
individual did not obtain as an advance can be obtained when he or she 
files a tax return, assuming qualifications are maintained. 

[3] GAO, Earned Income Tax Credit: Advance Option Is Not Widely Known 
or Understood by the Public, GAO/GGD-92-26 (Washington, D.C.: Feb. 19, 
1992). 

[4] U.S. Troop Readiness, Veteran's Care, Katrina Recovery, and Iraq 
Accountability Appropriations Act of 2007, Pub. L. No. 110-28. 

[5] 26 U.S.C. Sec. 32. 

[6] Pub. L. No.103-66 (1993). The Omnibus Budget Reconciliation Act of 
1993 limited the amount of advance payments that eligible individuals 
can receive to 60 percent of the EITC available with one qualifying 
child. 

[7] The Form W-5, completed by the employee, expires at the end of each 
calendar year. Therefore, employees must resubmit the form each year if 
they want to receive the AEITC. 

[8] When both spouses have a Form W-5 in effect, the amount each person 
can receive is reduced to below 60 percent of the EITC maximum. 

[9] Individuals who receive the EITC without a qualifying child are not 
eligible for the AEITC. 

[10] For the first notice mailing, specific information on how the 
population was selected, for example, is not known. 

[11] IRS, National Research Office, Advance Earned Income Tax Credit 
1994 and 1997 Notice Study, A Report to Congress (Washington, D.C.: 
August 1999). 

[12] The Small Business Taxpayer Education Program provides help to 
those who are starting or already have a small business and need 
information in such areas as taxes, recordkeeping, accounting 
practices, and completing federal business and employment tax returns. 
Much of the assistance is free. 

[13] We identified other studies with limited populations that also 
found taxpayers preferred receiving a lump sum instead of smaller 
periodic AEITC payments; 
for example, Jennifer Romich and Thomas Weisner, "How Families View and 
Use the EITC: Advance Payment vs. Lump Sum Delivery," National Tax 
Journal (December 2000). 

[14] One reason an AEITC recipient may not elect the advance in a 
subsequent year is because he or she no longer meets the eligibility 
requirements. 

[15] AEITC recipients must expect to be able to claim the EITC, which 
requires that taxpayers and their spouses, if filing a joint return, 
have valid SSNs issued by SSA to all U.S. citizens or individuals 
eligible to work. 

[16] If any portion of the invalid SSNs is attributable to aliens not 
currently authorized to work in the United States, the number of 
invalid SSNs could be affected if the Congress changes policies 
regarding such aliens. See Treasury Inspector General for Tax 
Administration, The Internal Revenue Service's Individual Taxpayer 
Identification Number Creates Significant Challenges for Tax 
Administration, 2004-30-023 (January 2004), for more information about 
aliens' use of invalid SSNs. 

[17] There could be some instances where the SSN reported on the Form W-
2 does not match the SSN the taxpayer used on the tax return. Even 
still, IRS uses the SSN on the Form W-2 and other third party forms for 
document matching to identify noncompliance such as wage 
underreporting. 

[18] Of the more than 100,000 AEITC recipients with an invalid SSN, 
most involved a SSN that did not match the individual's name on the 
Form W-2. 

[19] IRC § 6109(a)(1). 

[20] 42 U.S.C. § 408. 

[21] IRC §6723. 

[22] IRS can file a substitute return for people who do not voluntarily 
file. IRS assesses a tax liability and notifies the taxpayer of any tax 
due based on the substitute tax return, which is created from available 
Form W-2 and other information. 

[23] There could be instances where our methodology did not detect a 
filed tax return; 
for example, when a taxpayer's SSN on the Form W-2 might have been 
incorrect. See our scope and methodology (app. II). 

[24] Because AEITC is an advance payment of the EITC, filers who claim 
the EITC are required to reduce the EITC by any AEITC received. By 
examining the Form(s) W-2 of EITC claimants for AEITC payments, the IRS 
attempts to ensure that excess amounts of EITC are not claimed. 

[25] Submission Processing did not have tax year 2002 data available. 

[26] The electronic Form (s)W-2, which is an iteration of a user's 
entries into the tax preparation software, is generally included in the 
electronic filing submission. 

[27] ERS is the system for examining and correcting electronic and 
paper individual income tax returns rejected due to taxpayer and 
processing errors. ERS treats paper and electronic returns in the same 
way. 

[28] AUR categorizes a case as an AEITC case if AEITC is the primary 
issue, responsible for at least 80 percent of the increased tax 
liability. Thus, the total amount assessed in each AEITC case may 
include other underreported amounts. The total amount assessed is not 
necessarily the total amount that IRS collects from the taxpayer. 

[29] The automated collection system primarily focuses on collecting 
taxes from taxpayers who are delinquent in paying assessed taxes. 

[30] IRS officials said they decided to work on the cases for all 4 
years, even if they did not involve the taxpayer receiving the AEITC in 
some of those years so as to bring taxpayers in these cases into full 
compliance. 

[31] GAO, Earned Income Tax Credit: Implementation of Three New Tests 
Proceeded Smoothly, but Tests and Evaluation Plans Were Not Fully 
Documented, GAO-05-92 (Washington, D.C.: Dec. 30, 2004). 

[32] IRS does not track the number of AEITC cases that are audited. 

[33] IRS, Wage and Investment Research Group, Soft Notice for Duplicate 
TINS for Tax Years 2002, 2003, and 2004, 6-05-12-2-030E (Nov. 29, 
2005). 

[34] Booz Allen Hamilton, AUR Soft Notice Test Results (Oct. 6, 2005). 

[35] IRS, Wage and Investment Research Group, Dependent Database Soft 
Notice Report for Tax Year 2005, 6-06-12-2-0280E (Oct. 31, 2006). 

[36] There are different ways to calculate turnover. We considered it 
to be first time use within a 6-year period. To provide additional 
information, we also compared the number of first-time AEITC recipients 
who did not elect the advance in tax year 2003 with the total number of 
first-time AEITC recipients in tax year 2002. We made the same 
comparisons for the following year. 

[37] The Department of Homeland Security operates a voluntary program 
known as Employment Eligibility Verification, formerly known as Basic 
Pilot, through which participating employers enter employee 
information, such as name and SSN, into the Web site to verify newly 
hired employees' work authorization status. The program then attempts 
to match that information against SSA and, if necessary, Department of 
Homeland Security databases. We did not examine Employment Eligibility 
Verification as part of this report because we have identified several 
weaknesses in the program's implementation that could adversely impact 
increased use. See GAO, Immigration Enforcement: Weaknesses Hinder 
Employment Verification and Worksite Enforcement Efforts, GAO-06-895T 
(Washington, D.C.: June 19, 2006). 

[38] Financial institutions and other businesses that make certain 
payments must file a Form 1099 with IRS. Generally, these payments are 
not subject to withholding unless certain conditions exist, one of 
which is an invalid taxpayer identification number. 

[39] For the purposes of this report, SSN validity with regard to SSNVS 
includes instances where a name matches its SSN as verified by SSA 
records. 

[40] SSA, Office of Inspector General, Accuracy of the Social Security 
Administration's Numident File A-08-06-26100 (December 2006). When SSA 
assigns an SSN to an individual, the agency creates a master record in 
its Numident file, which contains information about the number holder, 
including the person's name. Thereafter, the Numident record for each 
number holder identifies any changes to the original information 
provided by the number holder, including name changes. 

[41] Supplemental Security Income is a federal program administered by 
SSA designed to provide cash for food, clothing, and shelter to aged, 
blind, and disabled people who have little or no income. 

[42] As of June 2007, this proposal was still outstanding. SSA sends 
"no match" letters to employers who submit more than 10 Forms W-2 to 
SSA or when more than 0.5 percent of these wage reports consist of a 
name and SSN combination that do not match SSA records. 

[43] Both IRS and SSA have toll-free telephone numbers listed on their 
Web sites that those looking for help with the particular service can 
call to speak with agency staff and get help. 

[44] The higher volume of questions in the first year assumes that 
employees whose name and SSN matched and who sought the AEITC from the 
same employer in a subsequent year would not be required to undergo a 
second match. 

[45] Internal Revenue Code Section 6103 prohibits the disclosure of 
information on tax returns, including SSNs, unless permitted by a 
specific exception. Because there is no exception for the verification 
of SSNs by employers, legislation would be needed to use the TIN 
Matching database for this purpose. 

[46] Office of Tax Policy, U.S. Department of the Treasury, Scope and 
Use of Taxpayer Confidentiality and Disclosure Provisions October 2000. 
Such a provision was included in S. 1321, 109 Cong. Sec. 509 (2006). 

[47] The Questionable W-4 program required employers to submit 
information to IRS on taxpayers who claimed more than 10 withholding 
allowances or exemptions. 

[48] GAO, IRS's Questionable Form W-4 Program, GAO-04-79R (Washington, 
D.C.: Nov. 6, 2003). 

[49] The number includes AEITC amounts received by nonfilers, filers 
who misreported the amount received, and individuals with invalid SSNs. 

[50] This estimate is illustrative only and many variables could affect 
how successful any compliance effort would be and the cost of 
implementing an effort. 

[51] The AEITC amount on the Form W-2 is what an employer reported 
paying to an employee. While this is not necessarily evidence that the 
employee actually received that amount, IRS considers the Form W-2 the 
official wage and tax statement and makes adjustments to the tax return 
based on the Form W-2 when discrepancies exist. 

[52] An individual who had more than one Form W-2 reporting AEITC for a 
particular year was counted as only one individual and the amount of 
AEITC paid was totaled. 

[53] The National Account Profile incorporates the DM-1 from SSA, which 
is the primary source file for all SSNs. The National Account Profile 
also includes additional SSN change information available to IRS to 
show the most current and complete taxpayer information. 

[54] The AEITC maximum limits for tax years 1999-2004 were: $1,387, 
$1,412, $1,457, $1,503, $1,528, $1,563, respectively. These amounts are 
limited if two married individuals both elect the AEITC. 

[55] While each subpopulation contains a unique set of Form W-2s, 
during tax years 2002-2004 there were 856 Form W-2s that appear in both 
the valid and invalid name subpopulations. This could occur due to 
variations in how the name is spelled or reported by the employer and/ 
or due to timing differences in reporting and updating name changes to 
SSA. 

[56] Because the SSN was invalid, it is not possible to determine 
whether the individual's name matched. 

[57] The Compliance Data Warehouse houses excerpts of data from various 
IRS systems for easy retrieval. 

[58] The EITC Database is an extraction of EITC data from IRS's 
Individual Returns Transaction File. The database is located on the 
Compliance Data Warehouse. These data met our reliability standards. 

[59] We spoke with Leonard Burman, Urban Institute; 
Leslie Book, Villanova University School of Law; 
Jonathan Forman, University of Oklahoma College of Law; 
William Gale, The Brookings Institution; 
Robert Greenstein, Center on Budget and Policy Priorities; 
Teresa Hinze, Community Tax Aid, Inc; 
Julie Kruse, Center for Economic Progress; 
Diana Leyden, University of Connecticut School of Law; 
David Marzahl, Center for Economic Progress; 
John Karl Scholz, University of Wisconsin Department of Economics; 
and Eugene Steuerle, Urban Institute. 

[60] The large difference in wages between the AEITC valid and invalid 
name subpopulations could be a result of women who married during the 
year and changed their name with their employer, but not SSA. These 
women likely reported combined wages with their spouse. 

[61] GAO/GGD-92-26 and Treasury Inspector General for Tax 
Administration, Taxpayers Were Assessed Additional Tax for Advance 
Earned Income Credit Payments Not Received, 2003-40-126 (June 2003). 

[62] We only included laws that made specific changes to the AEITC or 
its administration that were independent of changes to the EITC. Thus, 
we did not include laws whose only change to the AEITC occurred because 
of changes to the broader EITC program, such as changes in EITC benefit 
amounts or eligibility rules. 

[63] H.R. Conf. Rep. No. 103-213 (1993). 

[64] Pub. L. No. 110-28. 

GAO's Mission: 

The Government Accountability Office, the audit, evaluation and 
investigative arm of Congress, exists to support Congress in meeting 
its constitutional responsibilities and to help improve the performance 
and accountability of the federal government for the American people. 
GAO examines the use of public funds; evaluates federal programs and 
policies; and provides analyses, recommendations, and other assistance 
to help Congress make informed oversight, policy, and funding 
decisions. GAO's commitment to good government is reflected in its core 
values of accountability, integrity, and reliability. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each 
weekday, GAO posts newly released reports, testimony, and 
correspondence on its Web site. To have GAO e-mail you a list of newly 
posted products every afternoon, go to [hyperlink, http://www.gao.gov] 
and select "Subscribe to Updates." 

Order by Mail or Phone: 

The first copy of each printed report is free. Additional copies are $2 
each. A check or money order should be made out to the Superintendent 
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or 
more copies mailed to a single address are discounted 25 percent. 
Orders should be sent to: 

U.S. Government Accountability Office 441 G Street NW, Room LM: 
Washington, D.C. 20548: 

To order by Phone: Voice: (202) 512-6000: 
TDD: (202) 512-2537: 
Fax: (202) 512-6061: 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 

Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470: 

Congressional Relations: 

Gloria Jarmon, Managing Director, JarmonG@gao.gov (202) 512-4400: 

U.S. Government Accountability Office, 441 G Street NW, Room 7125: 
Washington, D.C. 20548: 

Public Affairs: 

Susan Becker, Acting Manager, Beckers@gao.gov (202) 512-4800: 

U.S. Government Accountability Office, 441 G Street NW, Room 7149: 
Washington, D.C. 20548: