Earned Income Credit: Targeting to the Working Poor (Briefing Report,
03/31/95, GAO/GGD-95-122BR).

Pursuant to a congressional request, GAO: (1) provided information on
Earned Income Credit (EIC) noncompliance and the Internal Revenue
Service's (IRS) actions to control EIC noncompliance; (2) reviewed the
potential impacts of changes in EIC eligibility criteria; and (3)
provided information on how many illegal aliens receive EIC and the
administration's proposal to exclude illegal aliens from EIC.

GAO found that: (1) reliable IRS noncompliance measures for the EIC
program do not exist and reliable data on the current extent of
noncompliance is unavailable; (2) in 1988, about 42 percent of
recipients received too much EIC and about 32 percent could not document
their EIC entitlement; (3) in a sample of 1994 electronic returns, 29
percent of these recipients received too much EIC and 13 percent
deliberately claimed too much EIC; (4) IRS actions to detect erroneous
EIC claims include developing improved criteria for detecting multiple
use of the same social security number (SSN) and valid SSN for
qualifying children; (5) adding taxpayers' wealth and other income
sources as EIC eligibility criteria could produce $318 million to a few
billion dollars in EIC savings; (6) the additional EIC eligibility
criteria would increase EIC complexity and the administrative burden of
collecting needed data and treating similar taxpayers equally; (7) IRS
does not know how many illegal aliens receive EIC, since it does not
prohibit them from claiming EIC or require them to identify themselves
on IRS forms; (8) administration and legislative proposals would exclude
illegal aliens from EIC eligibility; and (9) the administration's
proposal would require all EIC claimants to have valid work-related SSNs
and permit streamlined IRS procedures to enforce the requirement.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  GGD-95-122BR
     TITLE:  Earned Income Credit: Targeting to the Working Poor
      DATE:  03/31/95
   SUBJECT:  Tax administration
             Noncompliance
             Income taxes
             Tax credit
             Illegal aliens
             Fraud
             Erroneous payments
             Eligibility criteria
             Disadvantaged persons
             Statistical methods
IDENTIFIER:  Earned Income Tax Credit
             IRS Taxpayer Compliance Measurement Program
             IRS Earned Income Credit Compliance Initiative
             Tax Compliance Act of 1995
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* No attempt has been made to display graphic images, although figure    *
* captions are reproduced. Tables are included, but may not resemble     *
* those in the printed version.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Briefing Report to the Chairman, Committee on Governmental Affairs,
U.S.  Senate

March 1995

EARNED INCOME CREDIT - TARGETING
TO THE WORKING POOR

GAO/GGD-95-122BR

Earned Income Credit


Abbreviations
=============================================================== ABBREV

  AFDC - Aid to Families With Dependent Children
  AGI - Adjusted Gross Income
  CPS - Current Population Survey
  EIC - Earned Income Credit
  IRS - Internal Revenue Service
  JCT - Joint Committee on Taxation
  OBRA - Omnibus Budget Reconciliation Act
  SOI - Statistics of Income
  SSA - Social Security Administration
  SSN - Social Security number
  TIN - taxpayer identification number

Letter
=============================================================== LETTER


B-260826

March 31, 1995

The Honorable William V.  Roth, Jr.
Chairman, Committee on Governmental Affairs
United States Senate

Dear Mr.  Chairman: 

This is our second report responding to your interest in the Earned
Income Credit (EIC).\1 In this report, which follows our briefing for
you, we present information on EIC noncompliance and assess changes
and administrative issues that might result from potential changes to
EIC eligibility criteria.  These criteria would take into account (1)
measures of taxpayer wealth and (2) more sources of income when
determining who qualifies for the credit.\2 We also provide
information on illegal alien recipients of the EIC and describe the
administration's proposal, which is similar to your 1994 proposal, to
exclude such aliens from eligibility. 


--------------------
\1 Tax Administration:  Earned Income Credit--Data on Noncompliance
and Illegal Alien Recipients (GAO/GGD-95-27, Oct.  25, 1994) focused
on EIC noncompliance. 

\2 As you requested, we initially assessed the magnitude of change
likely to result from taking wealth and additional sources of income
into account when awarding the EIC.  On the basis of this work, you
requested that the Joint Committee on Taxation (JCT) provide official
revenue estimates.  We present those estimates in this report. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

EIC noncompliance has been a continuing concern of the Internal
Revenue Service (IRS).  Current, reliable noncompliance measures do
not exist for the entire EIC program.  An IRS study of noncompliance
for returns filed electronically during two weeks in January 1994
found that an estimated 29 percent of these recipients received too
much EIC, and 13 percent intentionally claimed too much.  The extent
of such noncompliance for paper returns is unknown but also of
concern to the IRS. 

Concerned about EIC noncompliance and refund fraud generally, IRS has
taken steps to detect and prevent erroneous refund payments.  These
include developing and applying improved criteria for detecting
noncompliant returns and checking for the use of the same Social
Security number (SSN) on multiple tax returns.  These steps have
resulted in many more taxpayers being asked to provide evidence of
EIC eligibility and in delaying refunds to at least 2.9 million EIC
claimants as of March 17, 1995.  In addition, as of March 17, IRS had
sent out almost 4.1 million notices primarily when returns did not
appear to contain valid SSNs for dependents or, in the case of EIC,
for qualifying children.  Although these steps may inconvenience or
burden taxpayers, if implemented effectively, they could help IRS
improve EIC compliance. 

Taxpayers' earned income, and in some cases their adjusted gross
income (AGI), as well as whether they have children meeting certain
age and residency tests, determine EIC eligibility and credit
amounts.  Unlike certain federal welfare programs, taxpayers' wealth
(e.g., the value of property or other investments they own) does not
affect EIC eligibility.  In addition, the EIC does not consider
certain forms of income in determining how much, if any, credit
taxpayers will receive.  EIC eligibility criteria could be changed to
take into account wealth and additional forms of income. 

The JCT estimates that denying the EIC to taxpayers who have some
wealth, as indirectly measured by their asset-derived income, could
yield $318 to $971 million in revenue savings in fiscal year 1997,
depending on the wealth test design.  These revenue savings represent
potential reductions in EIC program costs resulting from changing EIC
eligibility criteria.  In addition, taking nontaxed Social Security
income, tax-exempt interest, and nontaxed pension distributions into
account in taxpayers' AGI for credit calculations could yield $1.449
billion in revenue savings in fiscal year 1997, according to JCT
estimates.  Also, taking child support payments into account would
increase revenues in fiscal year 1997 by $686 million. 

However, adding an indirect wealth test or an expanded AGI definition
to the EIC eligibility criteria would add to the EIC's complexity. 
Complexity has been a continuing EIC issue because it can lead to
increased errors and dissuade deserving taxpayers from claiming the
credit.  Of the potential changes to EIC criteria, adding child
support payments to taxpayers' AGI likely would cause the greatest
complexity because information on such income is not collected by IRS
and systems may not exist to comprehensively generate the
information. 

Although an indirect wealth test for the EIC that uses tax return
data might be more practical than a more comprehensive test, it would
have significant limitations in measuring potential EIC recipients'
actual wealth.  For instance, such a test would not measure the value
of taxpayer assets like capital stock funds that yield little, if
any, annual income.  These limitations could raise concerns that
taxpayers with similar wealth would be treated differently for the
EIC. 


   BACKGROUND
------------------------------------------------------------ Letter :2

The EIC is a refundable tax credit available to low-income working
taxpayers with children and, beginning in tax year 1994, certain
taxpayers without children.  Congress established the EIC in 1975 to
achieve two long-term objectives:  (1) to offset the impact of Social
Security taxes on low-income workers with families and (2) to
encourage low-income individuals with families to seek employment
rather than welfare. 

For tax year 1993, about 14.7 million taxpayers claimed about $15
billion in EIC benefits.  To be eligible for any EIC in tax year
1993, a taxpayer must have had earned income of less than $23,050 and
had one or more qualifying children who met the age, relationship,
and residency tests.  The Omnibus Budget Reconciliation Act (OBRA) of
1993 increased the number of taxpayers eligible for the EIC and the
credit amount.  These changes began in tax year 1994 and will be
fully effective in tax year 1996.  The maximum income qualifying for
the EIC will rise to $27,000 in tax year 1996, the maximum credit
will rise to $3,370 for tax year 1996,\3 and total EIC cost in fiscal
year 1996 is expected to reach nearly $25 billion (in 1994 dollars). 


--------------------
\3 All monetary figures are in 1994 dollars.  Under OBRA 1993,
amounts would be higher in 1995 and 1996 than shown here because of
annual inflation adjustments. 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
------------------------------------------------------------ Letter :3

Our objectives were to (1) present information about EIC
noncompliance and what steps IRS is taking to control such
noncompliance and (2) review the impact on the amount of EIC paid
that might result from potential changes to the EIC eligibility
criteria that would reflect taxpayer wealth and additional sources of
income and administrative issues which could arise due to these
changes.  In addition, we were asked to provide information about
illegal aliens receiving the EIC and to describe the administration's
proposal to exclude illegal aliens from eligibility. 

To review the effects of possible changes to EIC eligibility
criteria, we obtained and analyzed data from the IRS' Statistics of
Income Division (SOI) and from the Bureau of the Census' Current
Population Survey (CPS).  For our various objectives, we also met
with Treasury, IRS, JCT, and Congressional Budget Office officials;
visited IRS Service Centers in Cincinnati, OH, and Fresno, CA; and
reviewed relevant literature on the EIC.  See appendix I for more
details on our methodology. 

We did our work from August 1994 through February 1995 in accordance
with generally accepted government auditing standards.  On March 21,
22, and 23, 1995, we discussed our draft report with Department of
the Treasury and IRS officials who are responsible for administering
the EIC, ensuring compliance, and analyzing potential policy changes. 
The officials generally agreed with the material in the report but
offered updated data and suggestions for improving the clarity of
presentation.  We made appropriate changes to the report to reflect
their comments. 


---------------------------------------------------------- Letter :3.1

As agreed with your office, unless you publicly announce the contents
of this report earlier, we will not distribute this report until
April 4, 1995.  At that time, we will send copies of this report to
various interested congressional committees, the Secretary of the
Treasury, the Commissioner of Internal Revenue, and other interested
parties.  We will also make copies available to others on request. 

The major contributors to this report are listed in appendix II. 
Please contact me on (202) 512-8633 if you have any questions about
this report. 

Sincerely yours,

Lynda D.  Willis
Associate Director, Tax Policy and
 Administration Issues


Briefing Section I BACKGROUND
============================================================== Letter 

Targeting the EIC


      ENSURING THAT THE WORKING
      POOR RECEIVE THE EIC
---------------------------------------------------------- Letter :3.2

As requested by the Chairman, Senate Committee on Governmental
Affairs, this report addresses how the federal government can better
ensure that only the working poor receive the EIC.  Specifically, the
report

presents information about noncompliance with the EIC and what the
IRS is doing to increase compliance and thus exclude ineligible
taxpayers from receiving the credit;

assesses changes that may result from potential changes to the
criteria used in determining EIC eligibility.  These changes would
take into account more of the resources that taxpayers could use to
support themselves and their families (resources not taken into
account when determining EIC eligibility include taxpayers' wealth
and certain forms of income);\4 and

presents information about how many illegal aliens receive the EIC
and discusses the administration's proposal, which is similar to
Senator Roth's 1994 proposal, to exclude illegal

Range of EIC Recipients With Two Qualifying Children (1996)



   (See figure in printed
   edition.)

   Source:  Congressional Research
   Service

   (See figure in printed
   edition.)


--------------------
\4 EIC recipients' wealth would include the value of assets like
savings, stock or property that they may own.  Additional income
sources not taken into account in determining how much, if any, EIC
to award include, for example, nontaxed Social Security income,
tax-exempt interest income, and nontaxed pension distributions. 


      HOW THE EIC IS AWARDED
---------------------------------------------------------- Letter :3.3

The EIC is a refundable tax credit available to low-income working
taxpayers with children and, beginning in 1994, certain taxpayers
without children.  Congress established the EIC in 1975 to achieve
two long-term objectives:  (1) to offset the impact of Social
Security taxes on low-income workers with families and (2) to
encourage low-income individuals with families to seek employment
rather than welfare. 

EIC eligibility and credit amounts generally are determined according
to the taxpayers' earned income and whether they have qualifying
children who meet age, relationship, and residency tests.  The credit
gradually phases in, plateaus at a maximum amount, and then phases
out until it reaches zero.  If the taxpayers' earned income or
adjusted gross income (AGI) exceeds the maximum qualifying income
level, they are not eligible for the credit.  When the taxpayers' AGI
falls in the credit's phase-out range, they receive the lesser amount
resulting from using their earned income or AGI in calculating the
credit. 

As the figure illustrates, when changes made in the 1993 Omnibus
Budget Reconciliation Act (OBRA) are fully in effect in tax year
1996, taxpayers with two children and whose earned income ranges from
$1 to $8,424 will receive $.40 for each dollar earned.  For taxpayers
with incomes between $8,425 to $10,999, the amount of EIC received
will remain stable at $3,370.  Taxpayers whose income falls between
$11,000 and $27,000 will receive a declining amount of EIC, with the
credit falling $.21 for each additional

Growth in EIC Program Costs (1988-2000)



   (See figure in printed
   edition.)

   Source:  Fiscal year estimates
   from the Presidents'

   (See figure in printed
   edition.)


      BROADER COVERAGE AND LARGER
      CREDIT AMOUNTS INCREASE EIC
      PROGRAM COSTS
---------------------------------------------------------- Letter :3.4

Total program costs\5 for the EIC have increased dramatically as
Congress has broadened its coverage and increased the amount of
credit available.  The figure shows that between 1988 and 1996 total
EIC program costs are estimated to increase over five fold in real
terms, from $4.4 billion in 1988 to an estimated $24.6 billion in
1996.  Congress has increased the coverage and amount of the credit
for reasons such as to (1) ensure that EIC amounts would not fall in
terms of purchasing power, (2) increase or maintain the progressivity
of the tax system, and (3) better ensure that working individuals
will have incomes above the poverty line. 

The most recent change to the EIC, in the OBRA of 1993, increased the
maximum credit available and the income level at which individuals
can qualify for the credit, and made certain low-income taxpayers
without children eligible.  The maximum credit amount for a family
with two children is rising from $1,511 for tax year 1993 to $3,370
in tax year 1996.  The maximum income qualifying for the EIC is
rising from $23,050 in tax year 1993 to $27,000 in tax year 1996. 
Finally, beginning in tax year 1994, individuals without a qualifying
child are eligible for the credit if they (1) are at least 25 but
less than 65 years old, (2) are not a dependent of another taxpayer,
and (3) have earned income and AGI of $9,000 or less.  These
taxpayers will be eligible for a maximum credit of $306, adjusted for
inflation. 

Briefing Section II

--------------------
\5 The EIC is a refundable tax credit.  As such, the portion of the
credit that offsets taxes owed by EIC recipients is considered a tax
expenditure, i.e., a reduction in taxes due to a preferential
provision in the federal tax law.  The refundable portion of the EIC
is considered a federal outlay.  We totaled the tax expenditure
estimate and the outlay estimate from appropriate versions of the
President's Budget to arrive at "total EIC program costs."


NONCOMPLIANCE
============================================================== Letter 

IRS Studies



   (See figure in printed
   edition.)

   Source:  IRS data.

   (See figure in printed
   edition.)


      IRS IS STUDYING EIC
      NONCOMPLIANCE
---------------------------------------------------------- Letter :3.5

Currently, no reliable data exist on the extent of noncompliance
among all EIC claimants.  The most recent Taxpayer Compliance
Measurement Program, from 1988, showed that about 42 percent of EIC
recipients received too large a credit and about 32 percent were not
able to show they were entitled to any credit.  About $1.9 billion
(34 percent of the total EIC paid out) was awarded erroneously. 
However, the impact of the significant changes to the EIC since 1988
suggest that a new compliance measurement is needed.  For instance,
in the intervening years, Congress has broadened the EIC coverage,
increased the credit amount, and revised filing status and qualifying
children criteria. 

In response to concerns about EIC-related fraud, IRS studied a sample
of those returns filed during a 2-week period in January 1994.  Study
results are only generalizable to electronic returns filed during
this period.  IRS' preliminary analysis of these returns showed that
an estimated 29 percent of the 1.3 million EIC returns filed
electronically during the period had claimed too large a refund and
about 13 percent of the returns filed were estimated by IRS as having
intentionally claimed too much EIC.  Of the $1.5 billion of EIC
claimed in this period, an estimated $358 million was erroneously
claimed--about $183 million, or 12 percent, was classified as
intentional error.  This intentional error category comes closest of
any IRS category in the study to measuring EIC fraud.\6 About 3
percent of taxpayers claimed a total of about $7 million less EIC
than they were entitled to receive. 

In the fall of 1994, IRS began reanalyzing the electronic returns
using additional income data that were not available earlier.  The
results of the additional analyses are not yet available.  IRS
officials expect the analyses to lead to a higher estimated error
rate. 

IRS is doing a 1995 study that will yield a noncompliance estimate
for the entire EIC program.  The study will include a random sample
of EIC returns filed electronically and on paper throughout the 1995
filing season.  Preliminary results may be

1994 Filing Season Enforcement Results



   (See figure in printed
   edition.)

   Source:  IRS data.

   (See figure in printed
   edition.)


--------------------
\6 Determining whether a refund is fraudulent requires determining
the taxpayer's intent, which is difficult to prove. 


      IRS' 1994 ENFORCEMENT
      EFFORTS ADDRESS
      NONCOMPLIANCE
---------------------------------------------------------- Letter :3.6

In addition to the 2-week electronic EIC return study, IRS detected
noncompliance in its normal efforts to detect inaccurate or improper
returns as they are processed.  These data document noncompliance
that IRS discovers while processing EIC returns but do not measure
the universe of noncompliance. 

In 1994, during initial manual reviews of tax year 1993 EIC paper
returns, IRS personnel identified about 150,000 taxpayers who claimed
the EIC although the return information indicated they were not
entitled to it.  Most problems resulting in disqualification related
to qualifying children, such as a child exceeding the age limit. 
IRS' initial computerized reviews of about 6 million electronically
filed EIC returns resulted in about 610,000 rejection notices being
sent out.  The rejections occurred when the qualifying child's Social
Security number (SSN) did not match the Social Security
Administration's (SSA) records. 

Beginning in January 1994, IRS personnel also stopped the processing
of returns that lacked an SSN for the qualifying child and had a tax
refund exceeding a threshold.  Following statutorily required notice
procedures, IRS suspended the EIC refund and asked the taxpayer to
submit proof of their EIC qualification.  If proof was provided, the
refund was released.  If the taxpayer submitted insufficient proof or
failed to respond, IRS' policy was to permanently deny the refund. 
As of September 30, 1994, IRS had delayed about $500 million in
potentially erroneous EIC refunds claimed on about 400,000 of about
8.7 million paper returns.  IRS officials expected most of these
refunds to be permanently denied because many taxpayers did not
respond to requests for information or could not support their
claims. 

After electronic returns pass initial computer checks and paper
returns pass manual checks, data is entered into IRS computers, which
then identify returns that are potentially fraudulent.  These
potentially fraudulent returns are reviewed by fraud detection teams. 
The number of fraudulent returns detected has grown steadily over
recent years.  As of December 31, 1994, of the total number of
returns reviewed, IRS had identified 77,781 as fraudulent--44,137 on
paper and 33,644 electronic returns.  About $43 million in fraudulent
refunds was not detected soon enough to stop the refund to taxpayers,
but IRS stopped about $117 million in refunds from being released. 
About 91 percent of

1995 Interim Enforcement Results



   (See figure in printed
   edition.)

   Source:  IRS data.

   (See figure in printed
   edition.)


      IRS' 1995 COUNTERMEASURES
      FOR ADDRESSING NONCOMPLIANCE
---------------------------------------------------------- Letter :3.7

Verifying SSN accuracy is key to IRS' 1995 EIC enforcement efforts. 
For paper returns, IRS enters into computers the taxpayer's SSN and,
starting this year, dependent and EIC qualifying childrens' SSNs. 
When returns have missing or invalid SSNs (i.e., do not match SSA
records), IRS delays the return and contacts taxpayers to resolve the
problem.  As of March 17, 1995, IRS had delayed the refunds for about
355,000 paper returns that lacked a valid dependent or qualifying
child's SSN. 

IRS has added controls to prevent returns with missing or invalid
SSNs, or SSNs already used by another taxpayer, from being filed
electronically.  All returns with these problems are to be rejected
and returned for correction.  As of March 17, IRS had sent out 3.7
million rejection notices\7 principally for electronic returns with
SSN problems related to a questionable refund.  About 1.1 million
notices were primarily due to the EIC qualifying child's SSN or year
of birth not matching SSA records.  The remaining 2.6 million notices
were primarily due to dependent SSN problems. 

In addition, IRS is working to identify uses of the same SSN on more
than one tax return.  Through a new tracking system, IRS intends to
identify potentially problematic returns and trigger enforcement. 
Because of past EIC fraud problems, IRS is concentrating on EIC
returns.  IRS delays refunds up to 8 weeks from the time a notice is
sent to taxpayers to allow staff time to identify duplicate SSN uses
and fraud schemes--about 7 million EIC returns could be delayed.  As
of March 17, about 2.9 million EIC refunds had been delayed.  As of
March 17, IRS national office officials told us that initial problems
with the duplicate SSN system were overcome early in the year. 
However, compliance personnel we spoke with said that problems with
duplicate SSN data continued to impede their effective use of the
system in mid-March. 

IRS has begun checking criminal and credit histories of new return
originator applicants who wish to file taxpayers' returns
electronically.  Due to these checks, IRS had rejected about 1
percent of applicants. 

As of mid-March, statistics were not available on fraudulent returns
detected this year. 

Briefing Section III

--------------------
\7 Because a return can be rejected for more than one reason, the
number of notices may be exceed the number of returns. 


BETTER MEASURING EIC FILERS'
RESOURCES
============================================================== Letter 

Wealth



   (See figure in printed
   edition.)


      OPTIONS FOR MEASURING WEALTH
---------------------------------------------------------- Letter :3.8

Congress requires income and wealth tests for certain welfare
programs like Aid to Families With Dependent Children (AFDC).  States
use questionnaires to determine the welfare applicant's degree of
need.  A similar wealth test for the EIC likely would require
additional IRS resources, or a diversion of current resources, to
obtain and verify the data.  Although states might administer a
wealth test for IRS, such an arrangement likely would take time to
perfect. 

Alternatively, a test that uses income earned from assets as an
indirect indicator of wealth is perhaps more immediately practical.\8
Such a test could measure income reported on tax returns that is
derived from taxpayers' assets and compare that income to an income
threshold.  This is the general approach proposed by the
administration and incorporated in House and Senate versions of H.R. 
831, a bill to permanently extend the health care deduction for
self-employed individuals. 

In evaluating an indirect wealth test, Congress might wish to
consider several options.  Asset-derived income is in several income
categories.  These include taxable interest and dividends, tax-exempt
interest, estate and trust income, rental income, and capital gains. 
A wealth test that includes a broad array of asset-derived income
might better measure taxpayers' wealth than a less inclusive test. 

However, no wealth test relying on tax-return information can
completely measure a taxpayer's wealth.  For example, the value of a
home,\9 valuable collections, and stocks that appreciate but pay few,
if any, dividends is not reflected on tax forms except when the
assets are sold.  Within the constraints of using data reported to
IRS, broader income measures might come closest to measuring a
taxpayer's overall wealth, but the measures nevertheless could
incorrectly represent some taxpayers' wealth while more accurately
measuring others' wealth.  These limitations raise fairness concerns
since taxpayers with similar

Wealth Thresholds



   (See figure in printed
   edition.)

   \ a The relationship between
   income and wealth is based on
   the assumption of a simple
   6-percent annual realized
   return on the value of the
   underlying assets.

   (See figure in printed
   edition.)


--------------------
\8 Taxpayers' wealth is somewhat taken into account through the
present AGI rule since AGI includes some asset-derived income.  The
wealth test we discuss takes this approach an additional step by
disqualifying taxpayers' whose asset-derived income, when summed,
exceeds a threshold. 

\9 The value of a recipient's home may not be included in wealth
tests for welfare programs; for example, it is not considered for
AFDC. 


      WEALTH THRESHOLDS FOR
      QUALIFYING FOR THE EIC
---------------------------------------------------------- Letter :3.9

Congress will need to set a threshold amount above which taxpayers
would be disqualified from receiving the EIC if it wishes to adopt an
indirect wealth test for the credit.  The House version of H.R.  831
proposes a $2,500 threshold, indexed for inflation.  The credit would
phase out as asset-based income rose above $2,500 and disappear when
such income was at least $3,150.  The Senate version of the bill
proposes a $2,450 unindexed threshold with different income items and
provides no phase out.  One consideration in selecting a threshold
amount could be the level of assets, or wealth, that a given level of
income may represent.  Possible thresholds could include, for
example, amounts of $1,000, $1,500, and $2,500.  Assuming a 6-percent
annual realized rate of return, about $16,700 of assets would
generate $1,000 of income.  About $25,000 and $41,700 of assets would
generate $1,500 and $2,500 of income, respectively. 

However, the relationship between asset-derived income and the
underlying asset value may vary widely.  For example, if a taxpayer
reports $1,000 of bank account interest, the average annual account
balance likely would have been between $10,000 (at a 10-percent
interest rate) and $30,000 (a 3.3-percent rate).  If $1,000 of
interest was earned on a tradeable bond, the value of the underlying
bond could lie outside those bounds because a bond's value rises or
falls as interest rates change. 

Associating income reported on tax forms with an underlying asset
value is most problematic for capital gains income.  For example,
$1,000 of reported gain could come from a successful $1,000
investment in stocks that doubled in value.  But, a $1,000 capital
gain also could come from a $100,000 investment in stock that
performed very poorly.  The association between reported capital
gains and underlying asset values also is complicated because the
return on the assets could have been accumulated over many years. 

It is difficult to say whether a goal of treating taxpayers of
similar means similarly is better served by implementing a broad
wealth test that combines income from assets with widely varying
rates of return, or implementing a narrower test that ignores some
assets completely.  In considering a threshold amount for an indirect
wealth test, lower thresholds may be more appropriate if the
relationship between income and the value of underlying

Wealth Test Results



   (See figure in printed
   edition.)

   \ a Less investment interest
   paid.

   (See figure in printed
   edition.)

   \ b Net rental real estate
   income, net income from Real
   Estate Mortgage Investment
   Conduits, net farm rental
   income, and royalty income.

   (See figure in printed
   edition.)

   \ c Net capital gains (Schedule
   D) and other gains (Schedule
   4797).

   (See figure in printed
   edition.)


      LOWER WEALTH THRESHOLDS AND
      BROADER MEASURES REDUCE EIC
      PROGRAM COSTS
--------------------------------------------------------- Letter :3.10

As more sources of income derived from taxpayers' assets are included
in an EIC wealth test and as the test threshold is lowered, EIC
program costs would be reduced further.  Joint Committee on Taxation
(JCT) revenue estimates illustrate this point.\10 For example, option
3, including the broadest base of income derived from taxpayers'
assets, coupled with a $1,000 threshold, is estimated by JCT to raise
$971 million in revenue in fiscal year 1997.  In contrast, option 1,
including only taxable interest and dividends and tax-exempt
interest, coupled with the higher $2,500 threshold, is estimated to
yield $318 million in revenue in fiscal year 1997.  JCT's latest
published estimate for the EIC program's total cost in fiscal year
1997 is $25.8 billion.  Thus, $971 million would represent about a
3.8 percent reduction in the total EIC program costs, and $318

Income


--------------------
\10 Although figures provided by JCT are revenue estimates, these
estimates can be thought of as "program cost" reductions.  The EIC is
a refundable tax credit.  As such, the portion of the credit that
offsets taxes owed by EIC recipients is considered a tax expenditure,
i.e., a reduction in taxes due to a preferential provision in the
federal tax law.  The refundable portion of the EIC is considered a
federal outlay.  Since most EIC recipients owe no taxes, most of
total EIC "costs" come from the refundable, or outlay, portion of the
credit. 


      AGI CRITERIA COULD BE
      EXPANDED
--------------------------------------------------------- Letter :3.11

Although the amount of EIC a taxpayer receives is based largely on
earned income, the amount, if any, also depends in part on other
sources of income.  Taxpayers' AGI can limit their EIC payments.  In
addition to earned income, AGI includes income from other sources,
such as investments, alimony received, and unemployment compensation. 
When taxpayers' AGIs fall within the EIC phase-out range, EIC
payments are the lower of those resulting from using taxpayers' AGI
or earned income.  When AGIs exceed the top of the EIC phase-out
range, taxpayers are ineligible for the credit regardless of their
earned income level.  Adding income elements to calculations of the
AGI for the EIC, thus, would be an incremental change that would
enable Congress, if it so desired, to take into account a fuller
range of taxpayers' incomes in determining the amount of credit
taxpayers would receive.\11

One alternative for expanding AGI could be to include nontaxed Social
Security income, tax-exempt interest income, and nontaxed pension
income.  These income sources are excluded from AGI for purposes of
calculating income tax liabilities but are sources of support
available to individuals.\12 Of the three income items, Social
Security is the largest income source to EIC recipients.  Although
most taxpayers eligible for the EIC do not receive

Expanded AGI Results



   (See figure in printed
   edition.)


--------------------
\11 AGI would be expanded to include those other income items only
for purposes of the EIC and not for income tax liability. 

\12 The taxable portions of Social Security income and pensions are
included in taxpayers' AGIs. 


      RESULTS OF EXPANDING AGI TO
      INCLUDE NONTAXED SOCIAL
      SECURITY INCOME, TAX-EXEMPT
      INTEREST, AND NONTAXED
      PENSION DISTRIBUTIONS
--------------------------------------------------------- Letter :3.12

As requested by Senator Roth, JCT estimated the revenues that would
result from including nontaxed Social Security income, tax-exempt
interest income, and nontaxed pension income in taxpayers' AGI for
purposes of EIC eligibility.  It estimated that $1.449 billion in
revenue would be realized in fiscal year 1997 from this change.  This
would represent about a 5.6 percent reduction in the estimated $25.8
billion cost of the EIC program for fiscal year 1997. 

If both a wealth test and an expanded definition of taxpayers' AGIs
were adopted simultaneously for the EIC, the net result in revenues
would be somewhat less than the sum of savings from each test
independently.  This would occur because some of the taxpayers
disqualified by one test would also be disqualified by the other; but
these reductions resulting from the disqualified taxpayers should not
be counted twice when estimating net revenue

Additional Considerations



   (See figure in printed
   edition.)


      COMPLEXITY INCREASES WITH
      WEALTH AND INCOME TESTS
--------------------------------------------------------- Letter :3.13

A wealth test and an expanded AGI definition for credit determination
purposes would have additional consequences besides reducing EIC
program costs.  Both changes would increase the complexity of the EIC
and impose burdens on taxpayers in determining their eligibility and
on IRS in ensuring compliance.  Complexity has long been a concern of
the IRS.  Complexity contributes to taxpayer errors and the EIC's
high noncompliance rate.  IRS officials also expressed concern to us
that a wealth test might discourage some legitimately qualified
taxpayers from applying for the EIC--a longstanding concern.  Some
research suggests that between 14 percent and 25 percent of eligible
taxpayers do not claim the EIC.\13

Complexity concerns might be alleviated, in part, because a small
portion of EIC recipients have income that would be taken into
account in either the wealth test or the expanded AGI.  For example,
in 1992 about 82 percent of EIC recipients had no income from any of
the sources included in a broad indicator of wealth.  Furthermore,
about 50 percent of EIC recipients use paid preparers. 

Specifically for a wealth test, broader measures may capture more
sources of wealth and might reduce taxpayers' incentives to shift
investments to maintain their EIC eligibility.  However,IRS would be
unable to verify tax-exempt interest income because it receives no
third-party information reports to use in checking the accuracy of
taxpayers' returns.  Short of an audit, IRS also may be unable to
verify the cost basis used in determining certain capital gains that
could be included in a wealth test. 

For the expanded AGI definition, tax-exempt income would, of course,
be unverifiable.  In addition, only taxpayers whose Social Security
income is taxed report their Social Security income to IRS.  IRS
would have to collect Social Security income data for an expanded AGI
test.  The SSA provides all Social Security recipients with an annual
Form 1099 that records their Social Security income and a computer
tape containing this information is provided to IRS.  However, under
existing systems, according to the IRS, more than a year likely would
elapse before IRS would be able to match taxpayers' claimed Social
Security

Child Support Data Not Reported



   (See figure in printed
   edition.)

   Source:  JCT estimates for
   Senator William V.  Roth, Jr.

   (See figure in printed
   edition.)


--------------------
\13 Yin et al., Improving the Delivery of Benefits to the Working
Poor:  Proposals to Reform the Earned Income Tax Credit Program,
American Tax Policy Institute, Feb.  1994. 


      INCLUDING CHILD SUPPORT
      PAYMENTS IN AGI REDUCES
      COSTS BUT IMPOSES GREATER
      ADMINISTRATIVE DIFFICULTY
--------------------------------------------------------- Letter :3.14

The AGI definition for EIC purposes could be expanded to include
child support payments.  Including child support payments would
recognize that such payments are part of a family's support.  As
estimated by JCT, adding child support to AGI along with the other
items discussed earlier would result in a savings of $2.135 billion
in fiscal year 1997 (about a 8.3 percent reduction in total EIC
program costs for that year).  This is $686 million more than if the
AGI definition did not include child support payments. 

However, the administrative issues associated with incorporating
child support payments in EIC eligibility criteria appear to be more
formidable than for the other income items, such as Social Security
payments, that could be added to AGI for determining the EIC. 
Although child support payments may be a factor in determining
whether a divorced or separated parent may claim a child as a
dependent for income tax purposes, child support income itself is not
required on any IRS forms.  Therefore, if child support were to be
considered in determining EIC eligibility, IRS would need to begin
collecting this information. 

IRS' ability to ensure compliance is impeded when it cannot verify
the accuracy of information reported on tax returns.  Independent
verification of child support payments could be difficult.  In cases
where child support agreements are overseen by a court or a state or
local social services agency, the overseeing agency may be able to
report to the IRS the amounts of child support paid.  IRS could use
such a report to verify the accuracy of child support data used by
taxpayers claiming the EIC.  However, systems would need to be
developed to routinely report this data to IRS. 

When courts or social service agencies do not oversee child support
payments, third party verification of payments may be unavailable. 
Furthermore, the parents could have incentives to misreport the
amounts paid.  That is, a custodial parent might claim that payments
were not made because such payments could affect their ability to
claim the child as a dependent for tax purposes.  Noncustodial
parents who had not properly paid child support would have an
incentive to claim they had paid it if for no other reason than to
avoid child support enforcement procedures. 


Briefing Section IV EIC
ELIGIBILITY
============================================================== Letter 

Illegal Alien Recipients



   (See figure in printed
   edition.)


      SOME ILLEGAL ALIENS RECEIVE
      EIC
--------------------------------------------------------- Letter :3.15

The Internal Revenue Code does not prohibit illegal aliens from
receiving the EIC if they meet the prescribed eligibility
requirements.  IRS forms do not require illegal aliens to identify
themselves as such; therefore, IRS does not know how many illegal
aliens may be claiming and receiving the EIC. 

IRS needs an identification number, generally the taxpayer's SSN, to
process a tax return.  IRS assigns a temporary Taxpayer
Identification Number (TIN) when any taxpayer files a return with an
invalid SSN, a blank space, or the code "205(c)."\14 The designation
205(c) is often used by taxpayers to indicate they are not eligible
to receive an SSN.  Thus, IRS officials said taxpayers who enter this
code are likely to be illegal aliens. 

Limited data from manual reviews under the 1994 EIC Compliance
Initiative show that a minimum of 160,000 taxpayers,\15 out of about
8.7 million who filed paper returns claiming the EIC, entered 205(c)
instead of an SSN for a qualifying child.\16

Given use of the 205(c) code for qualifying children and their
enforcement experience, IRS officials believe these returns likely
were filed by illegal aliens.  IRS expects most of these refunds to
be denied because taxpayers will not be able to support their claims
by verifying that the dependent met the age, relationship, and
residency requirement. 

Some unknown portion of returns filed with SSNs may also be filed by
illegal aliens.  For example, if illegal aliens use SSNs of other
individuals when filing a return and IRS does not detect the SSN
duplication, they may receive an EIC refund.  IRS' new efforts to
detect duplicate uses of SSNs, if successfully implemented, should
reduce the number of illegal aliens as well

Excluding Illegal Aliens



   (See figure in printed
   edition.)


--------------------
\14 The designation 205(c) refers to section 205(c) of the Social
Security Act as amended which specifies the rules for issuing SSNs. 
Most legal aliens can obtain an SSN if they meet the applicable
requirements.  Section 205(c) does not refer specifically to illegal
aliens. 

\15 In addition to the 160,000, an unknown number of illegal aliens
would have received the EIC because the amount they claimed was below
the Compliance Initiative's dollar threshold. 

\16 EIC claimants are required to provide a TIN for qualifying
children. 


      PROPOSALS TO EXCLUDE ILLEGAL
      ALIENS FROM EIC ELIGIBILITY
--------------------------------------------------------- Letter :3.16

A Senate bill introduced in 1994 by Senator Roth and the
administration's Tax Compliance Act of 1995 (H.R.  981 and S.  453)
introduced in 1995 would deny the EIC to illegal aliens.  Illegal
aliens cannot be employed lawfully in the United States.  On the
other hand, the EIC, which is intended in part to encourage
employment, under current law, can be paid to illegal aliens.  Thus,
the EIC works at cross purposes with the prohibition on employment of
illegal aliens. 

The administration's proposal would require that all EIC recipients
provide SSNs that are valid for employment in the United States for
themselves, for their spouses, if applicable, and for qualifying
children.  Because illegal aliens cannot qualify for SSNs that are
valid for employment in the United States, they would not be able to
receive the EIC. 

The SSA provides the IRS with a computer tape with names and SSNs of
those individuals having SSNs.  The data include codes indicating
whether the SSN is for employment purposes or other purposes, such as
qualifying for Social Security benefits.  This data provides a tool
for IRS to use in enforcing the administration's proposal.  We have
not assessed the adequacy of the information in the tape for
enforcement purposes or the potential for illegal aliens to
fraudulently receive work-related SSNs from the SSA. 

Finally, the administration's proposal would permit IRS to use
streamlined procedures to enforce the requirement that EIC claimants
have valid work-related SSNs.  The administration proposes that IRS
be permitted to notify taxpayers who do not provide valid SSNs that
they are not eligible for the EIC.  Within 60 days, taxpayers would
either have to provide valid SSNs or request that IRS follow
deficiency procedures.  Deficiency procedures protect taxpayers'
rights through notices to the taxpayer and opportunities for rebuttal
of IRS' concerns and petition to the Tax Court.  Taxpayers who fail
to respond within 60 days to IRS' proposed notice regarding lack of
valid SSNs would be required to refile an amended return with correct
SSNs to obtain the EIC. 

The administration estimates that requiring all EIC recipients to
provide valid work-related SSNs and using streamlined procedures to
enforce this requirement would yield about $400 million in additional
revenue in fiscal year 1997. 


OBJECTIVES, SCOPE, AND METHODOLOGY
=========================================================== Appendix I

As requested, our objectives were to (1) present information about
EIC noncompliance and what steps IRS is taking to control
noncompliance and (2) review the impact on the amount of EIC paid
that might result from potential changes to the EIC eligibility
criteria that would reflect taxpayer wealth and additional sources of
income and administrative issues that could arise due to these
changes.  In addition, we were asked to provide information about
illegal aliens receiving the EIC and to describe the administration's
proposal, which is similar to Senator Roth's 1994 proposal, to
exclude illegal aliens from eligibility. 

In responding to all of these objectives, we met with and obtained
reports and data from officials with the Department of the Treasury
and IRS' national office.  Primarily in response to the objectives
concerning noncompliance and illegal aliens receiving the EIC, we
also met with officials and reviewed relevant procedures and data in
IRS' Cincinnati, OH, and Fresno, CA, Service Centers. 

Specifically for our objective of assessing the effects of certain
revisions to the criteria used in determining the amount, if any, of
EIC that is awarded, we

reviewed relevant literature on the EIC to understand its
requirements and to determine the results of others' analyses of EIC
eligibility modifications;

met with staff from the JCT to discuss advantages and disadvantages
of using possible proxy measures of wealth to determine EIC
eligibility;

interviewed Congressional Budget Office analysts regarding their
consideration of modifications to EIC eligibility requirements;

met with Department of the Treasury Office of Tax Analysis staff to
discuss policy and administrative issues associated with EIC
eligibility modifications; and

met with IRS national office officials, including the National
Director of Submissions Planning; the Chief of Service Center
Compliance; the Director of the Service's Tax Forms and Publications
Division; and the National Director, Applications Design and
Development Management, to discuss administrative issues associated
with EIC eligibility modifications. 

In addition, we performed statistical analyses to make preliminary
estimates of the potential effects on the overall amount of EIC
program costs and the number of EIC recipients of using measures of
taxpayers' wealth and additional sources of their income when
determining how much, if any, EIC would be awarded.  After we
determined the general magnitude of change that might result, we
discussed our preliminary results with Senator Roth and he requested
revenue estimates, shown in our study, from the JCT.\17

To do our analysis, we first determined how much EIC would have been
awarded and to how many recipients had the EIC rules that will be in
effect in 1996 actually been in effect in tax year 1992.  Applying
the 1996 rules, which increase the credit amount and the number of
individuals who are eligible for the EIC, provided us with a more
realistic indicator of the potential effects of changing the EIC
eligibility rules. 

Using this as a base measure, we computed the amount of EIC that
would have been awarded and the number of recipients if (1) wealth
measures of varying breadth coupled with varying cutoff thresholds
had been used in determining EIC eligibility and (2) nontaxed Social
Security income, nontaxed pension distributions, and tax-exempt
income had been included in taxpayers' AGIs in determining how much,
if any, EIC they would have received. 

To assess the likely effects of changing the EIC eligibility
criteria, we obtained and analyzed data from the Internal Revenue
Service's (IRS) Statistics of Income Division (SOI) and from the
Bureau of the Census' Current Population Survey (CPS).\18 The CPS
data was needed because SOI only has data that can be obtained from
tax returns.  Critically absent from SOI was data related to
nontaxable Social Security income.  (Social Security income is
reported on Forms 1040 and 1040A, but only to the extent that some
portion of the income is subject to tax.) We used tax year 1992 SOI
data and March 1993 CPS data, which incorporates economic data for
the calendar year 1992, because it conformed to the same period as
the most recent SOI data available. 

Using these data, we simulated the effect of broadening the
definition of AGI to include nontaxed Social Security income as well
as nontaxed private pensions and tax-exempt interest income.  (The
latter two items are reported on tax forms.) We used Census'
simulation of EIC recipients on its CPS data set to estimate the
Social Security income received by the actual SOI population. 
Overall, we believe that this simulation procedure yielded a
conservative estimate of the reductions in EIC program costs and the
number of affected EIC recipients. 


--------------------
\17 Our methodology produces an estimate given the taxpayer income
characteristics that existed in 1992 but assuming that the 1996 EIC
rules had then applied.  The result differs from a revenue estimate,
which provides an indication of future changes in revenues due to
changes in the EIC statute and also takes into account forecasted
changes in interest rates and other economic factors. 

\18 The Omnibus Budget Reconciliation Act of 1993 revised the EIC
criteria, in part making certain taxpayers without children eligible
for the credit.  Our analyses did not include taxpayers without
children. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix II

GENERAL GOVERNMENT DIVISION,
WASHINGTON, D.C. 

Michael Brostek, Assistant Director, Tax Policy and
 Administration Issues
James A.  Wozny, Chief Tax Economist, Tax Policy and
 Administration Issues
Mary Phillips, Assignment Manager
Clifford Tuck, Economist-in-Charge

CINCINNATI REGIONAL OFFICE

William Bricking, Evaluator-in-Charge
Ellen Soltow, Evaluator


GLOSSARY
=========================================================== Appendix 0

INCOME


      TAXABLE EARNED INCOME
------------------------------------------------------- Appendix 0:0.1

For the EIC taxable earned income includes (1) wages, salaries, and
tips, (2) union strike benefits, (3) long-term disability benefits
received prior to minimum retirement age, and (4) net earnings from
self-employment. 


      NONTAXABLE EARNED INCOME
------------------------------------------------------- Appendix 0:0.2

Among the earned income items that are nontaxable are (1) voluntary
salary deferrals, such as 401(k) plans or the federal thrift savings
plan, (2) pay earned in a combat zone, (3) basic quarter and
subsistence allowances from the U.S.  military, (4) housing allowance
or rental value of a parsonage for the clergy, and (5) excludable
dependent care benefits. 


      UNEARNED INCOME
------------------------------------------------------- Appendix 0:0.3

Items that are not earned income include (1) interest and dividends,
(2) Social Security and railroad retirement benefits, (3) welfare
benefits (including AFDC payments), (4) pensions or annuities, (5)
veterans' benefits, (6) workers' compensation benefits, (7) alimony,
(8) child support, (9) unemployment compensation (insurance), (10)
taxable scholarship or fellowship grants (not reported on Form W-2),
and (11) variable housing allowance for the military. 


      ADJUSTED GROSS INCOME
------------------------------------------------------- Appendix 0:0.4

In addition to taxpayers' earned income, AGI includes their income
from other sources, such as investments, alimony received, and
unemployment compensation. 

QUALIFYING CHILD

A qualifying child (1) is an EIC claimant's son, daughter, adopted
child, grandchild, stepchild, or foster child, (2) is under age 19 or
under age 24 and a full-time student or any age and permanently and
totally disabled, and (3) lives in the claimant's home in the United
States for more than half of the year (or all of the year if a foster
child). 

WORKING POOR

The term "working poor," while used in reference to the intended
beneficiaries of the EIC, is not defined in statute.  Generally, for
purposes of this report, we use the term to refer to those
individuals who meet the current EIC income criteria, or the revised
criteria that we analyze.  The revised criteria do not alter the
basic EIC income criteria, but rather include a fuller range of
potential EIC recipients' resources in determining whether basic
eligibility criteria are met. 

