Tax Policy: Information on the Joint and Several Liability Standard
(Letter Report, 03/12/97, GAO/GGD-97-34).
Pursuant to a legislative requirement, GAO reported on several issues
related to the joint and several liability standard that applies to
jointly filed federal income tax returns, focusing on: (1) the potential
universe of taxpayers that may be eligible for innocent spouse relief;
(2) the Internal Revenue Service's (IRS) practices and procedures for
handling requests for innocent spouse relief; (3) whether the innocent
spouse provisions provide the same treatment for all taxpayers; (4) the
potential effects of replacing the joint and several liability standard
with a proportionate liability standard; (5) the potential effects on
IRS of requiring it to abide by the terms of divorce decrees when those
decrees allocate tax liabilities; and (6) the potential effects on IRS
of changing the law so that community income of one spouse cannot be
seized to satisfy tax liabilities incurred by the other spouse before
their marriage.
GAO noted that: (1) it estimated that about 587,000 of the 48 million
couples who filed joint returns in 1992 had additional tax assessments
of more than $500; (2) this estimate represents the maximum number of
taxpayers potentially eligible for innocent spouse relief, however,
fewer would probably actually qualify; (3) although any taxpayer signing
a joint return may seek innocent spouse relief, according to IRS
officials, divorced taxpayers are more likely to face the most egregious
problems; (4) the limited information that was available indicated that
IRS received few requests for innocent spouse relief and denied most of
them; (5) GAO observed that IRS publications provide little information
on how to request innocent spouse relief and that the publications
covering procedures related to the need for relief have no information
on relief; (6) critics of the innocent spouse provisions contend that
the current provisions do not ensure that all deserving taxpayers
receive equivalent relief; (7) GAO estimated that for tax year 1992, an
additional 42,600 divorced taxpayers might have been eligible for
innocent spouse relief if the dollar thresholds had been eliminated; (8)
an alternative way to ensure that taxpayers are not held liable for
their spouses' taxes would be to replace the joint and several liability
standard with a proportionate liability standard; (9) under such a
standard, taxpayers would be responsible only for the taxes generated by
their individual incomes and assets or, for taxpayers living in
community property states, for the tax associated with one-half of the
community income; (10) divorcing couples may specify in their divorce
decrees how future liabilities resulting from their prior joint returns
are handled; (11) requiring IRS to be bound by divorce decrees is
impractical for two major reasons; (12) federal tax matters are the
exclusive jurisdiction of certain federal courts, while divorce matters
are generally handled by state courts; (13) IRS officials also raised
related concerns, such as whether their interpretation of lengthy and
complex divorce decrees would increase the number of appeals and whether
divorce decrees would be manipulated to reduce tax liabilities (14) IRS
can treat taxpayers living in community property states differently from
taxpayers living in common law states when collecting taxes; and (15)
since IRS does not maintain data on how often these levy actions occur,*
--------------------------- Indexing Terms -----------------------------
REPORTNUM: GGD-97-34
TITLE: Tax Policy: Information on the Joint and Several Liability
Standard
DATE: 03/12/97
SUBJECT: Personal liability (legal)
Tax returns
Tax nonpayment
Tax administration
Divorced persons
Tax return audits
Government publications
Taxpayers
Tax law
Eligibility criteria
IDENTIFIER: IRS Actuarial Audit Program
IRS Underreporter Program
IRS Audit Information Management System
IRS Problem Resolution Program
California
Iowa
Louisiana
Oklahoma
New York
Massachusetts
North Dakota
Wisconsin
IRS Integrated Collection System
******************************************************************
** 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. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
** A printed copy of this report may be obtained from the GAO **
** Document Distribution Center. For further details, please **
** send an e-mail message to: **
** **
** **
** **
** with the message 'info' in the body. **
******************************************************************
Cover
================================================================ COVER
Report to the Committee on Ways and Means, House of Representatives,
and the Committee on Finance, U.S. Senate
March 1997
TAX POLICY - INFORMATION ON THE
JOINT AND SEVERAL LIABILITY
STANDARD
GAO/GGD-97-34
Income Tax Liability
(268711)
Abbreviations
=============================================================== ABBREV
ABA - American Bar Association
AGI - Adjusted Gross Income
AICPA - American Institute of Certified Public Accountants
AIMS - Audit Information Management System
IRS - Internal Revenue Service
Letter
=============================================================== LETTER
B-271456
March 12, 1997
The Honorable William V. Roth, Jr.
Chairman, Committee on Finance
United States Senate
The Honorable Daniel P. Moynihan
Ranking Minority Member, Committee
on Finance
United States Senate
The Honorable Bill Archer
Chairman, Committee on Ways
and Means
House of Representatives
The Honorable Charles B. Rangel
Ranking Minority Member, Committee
on Ways and Means
House of Representatives
When a married couple files a joint federal income tax return, each
spouse becomes individually responsible for paying the entire amount
of the tax associated with that return. Because of this joint and
several liability standard, one spouse can be held liable for tax
deficiencies assessed after a joint return was filed that were solely
attributable to the actions of the other spouse. However, when one
spouse, independently and without the knowledge of the other spouse,
incurs the additional taxes, the other potentially "innocent spouse"
may obtain relief from the additional tax liability if certain
conditions are met. Because of concerns about the effectiveness of
the current innocent spouse provisions and other perceived inequities
caused by the joint and several liability standard, section 401 of
the Taxpayer Bill of Rights 2 directed us to study and report on
several issues related to the joint and several liability standard
that applies to jointly filed federal income tax returns.\1
Accordingly, this report discusses (1) the potential universe of
taxpayers that may be eligible for innocent spouse relief, (2) the
Internal Revenue Service's (IRS) practices and procedures for
handling requests for innocent spouse relief, (3) whether the
innocent spouse provisions provide the same treatment for all
taxpayers, (4) the potential effects of replacing the joint and
several liability standard with a proportionate liability standard,\2
(5) the potential effects on IRS of requiring it to abide by the
terms of divorce decrees when those decrees allocate tax liabilities,
and (6) the potential effects on IRS of changing the law so that
community income of one spouse cannot be seized to satisfy tax
liabilities incurred by the other spouse before their marriage.\3
--------------------
\1 Public Law 104-168, July 30, 1996. The law also directed the
Secretary of the Treasury to study and report on the same issues
discussed in this report. IRS and Treasury staff jointly conducted
the study. The report had not been issued at the time of our review.
\2 Under proportionate liability, each taxpayer is responsible only
for the taxes resulting from his or her individual income, even when
such income is reported on a joint return. Proportionate liability
was generally the standard followed before the Revenue Act of 1938,
which established joint and several liability on joint returns.
\3 In community property states, the income and assets of each spouse
belong equally to the other spouse and are available to pay the debts
(including taxes) of either spouse.
BACKGROUND
------------------------------------------------------------ Letter :1
A traditional argument for applying a joint and several liability
standard to joint returns is that married couples form a single
economic unit, so spouses who benefit from each other's income and
assets can be held responsible for the total tax liability generated
from the income and assets. The benefit of joint returns is that
married couples are taxed as if their combined income were equally
split between the spouses. For married couples with substantially
disproportionate incomes, such income-splitting may lower their
overall taxes because some of the higher earner's income could fall
into a lower tax bracket and be taxed at a lower rate than if it had
all been taxed as the income of one person.
The benefits of income-splitting first became available in community
property states in 1930 as a result of the Supreme Court case of Poe
v. Seaborn [282 U.S. 101 (1930)]. In that case, the Court held
that in community property states, the wife is vested with a
half-interest in her husband's income. Therefore, for federal tax
returns, income was divided equally between husband and wife,
regardless of who earned it. The benefits of income-splitting were
denied couples living in common law states as a result of another
1930 Supreme Court case, Lucas v. Earl [281 U.S. 111 (1930]. In
that case, the Supreme Court rejected a couple's private agreement
assigning one-half of each spouse's earnings to the other spouse for
federal income tax purposes. Income-splitting for all joint filers
was added to the Internal Revenue Code by the Revenue Act of 1948 as
a means of equalizing the tax rates for married couples in common law
states with the tax rates for those living in community property
states.
Congress subsequently determined that in some instances, it was
inequitable to hold taxpayers liable for additional taxes resulting
from their spouses' unreported income. A commonly cited example was
a spouse who, unknown to the other spouse, was engaged in an illegal
activity, did not report the illegal income, and was subsequently
caught and assessed the taxes on the illegal gain. In 1971, Congress
enacted the innocent spouse provisions in the Internal Revenue Code
(section 6013(e)) to recognize the inequity of holding spouses liable
for additional tax assessments in certain cases. The provisions were
broadened in 1984 to provide relief from liabilities resulting from
grossly erroneous deductions, credits, or basis (i.e., the purchase
price of an asset), in addition to unreported income. The current
innocent spouse provisions allow relief from the joint and several
liability standard when
-- the innocent spouse has filed a joint return with the culpable
spouse;
-- the spouse did not know and had no reason to know there was a
substantial tax understatement (knowledge test); and
-- taking into account all the facts and circumstances, it is
inequitable to hold the spouse liable for the additional tax
attributable to the substantial understatement of the culpable
spouse.
In addition, the spouse requesting relief must meet certain dollar
thresholds that vary depending on the cause of the additional
assessment:
-- A tax liability resulting from an omission of gross income must
exceed $500.
-- A tax liability resulting from a deduction, credit, or basis
that has no basis in fact or law must exceed $500 and also be in
excess of (1) 10 percent of the innocent spouse's adjusted gross
income for their preadjustment tax year if the taxpayer's income
is less than or equal to $20,000; or (2) 25 percent of the
innocent spouse's income if the taxpayer's income is greater
than $20,000. If the innocent spouse has remarried, the new
spouse's income is included in this calculation.
IRS generally follows state law in regard to ownership of property,
and the states define ownership of property very differently. In
community property states, the income and assets (property) of each
spouse belong equally to the other spouse and can be attached to pay
the debts (including taxes) of either spouse.\4 About 27 percent of
all taxpayers live in the nine community property states (Arizona,
California, Idaho, Louisiana, Nevada, New Mexico, Texas, Washington,
and Wisconsin). In common law states, spouses do not have an
inherent right to each other's income and assets. As a result, there
is a distinct difference in the application of joint and several
liability between the residents of community property states versus
common law states. For example, a married couple living in a common
law state can avoid joint and several liability by filing a "married,
filing separate" return, whereas filing a "married, filing separate"
return in a community property state will not necessarily result in a
similar avoidance of tax.
--------------------
\4 Community property principles apply to the income and assets
acquired during the marriage. Residents of community property states
may own separate property, which generally consists of property owned
before the marriage or property acquired by gift or inheritance. The
laws of the individual states govern the treatment of such separate
property and the income derived from it.
RESULTS IN BRIEF
------------------------------------------------------------ Letter :2
We estimated that about 587,000, or about 1 percent, of the 48
million couples who filed joint returns in 1992, the most recent year
for which data were available, had additional tax assessments of more
than $500. This estimate represents the maximum number of taxpayers
potentially eligible for innocent spouse relief; however, fewer would
probably actually qualify. Some taxpayers would also probably have
been assessed additional taxes as a result of overstated deductions,
credits, or basis, which have other dollar thresholds in addition to
the $500 threshold. Some of these taxpayers would also probably have
been ineligible for relief because they would not have met the
knowledge test of the innocent spouse provision. However, IRS did
not have the data that we would need to estimate the number of
taxpayers that fell into these categories. Furthermore, although any
taxpayer signing a joint return may seek innocent spouse relief,
according to IRS officials, divorced taxpayers are more likely to
face the most egregious problems because their current assets could
be seized to satisfy tax obligations accruing from joint returns
filed prior to the divorce. Assuming a 2 percent per year divorce
rate, we estimated that about 35,000 of the 587,000 couples with
additional tax assessments for 1992 were divorced in the 3 years
since the end of the 1992 tax year and thus a spouse from these
couples may have been more likely to consider applying for relief.
The limited information that was available from several operating
units within IRS indicated that IRS received few requests for
innocent spouse relief and denied most of them. Although information
was not available to determine why few requests were made, we
observed that IRS publications provide little information on how to
request innocent spouse relief and that the publications covering
procedures related to the need for relief (e.g., tax examination and
appeals) have no information on relief, and IRS has no specific form
or process for applying for innocent spouse relief.
Critics of the innocent spouse provisions contend that the current
provisions do not ensure that all deserving taxpayers receive
equivalent relief. For example, they believe that the dollar
thresholds for claiming innocent spouse relief may preclude some
deserving taxpayers from obtaining relief because of the amount of
their liability. Four states apparently agree because, for state
income tax purposes, they have no dollar thresholds. Assuming a 2
percent per year divorce rate, we estimated that for tax year 1992,
an additional 42,600 divorced taxpayers might have been eligible for
innocent spouse relief if the dollar thresholds had been eliminated.
Since the innocent spouse relief provisions relieve the innocent
spouse of responsibility for a joint liability rather than forgive
the tax obligation, we could not estimate the revenue impacts of
modifying the provisions because data were not available on how
successful IRS has been in collecting from culpable spouses.
An alternative way to ensure that taxpayers are not held liable for
their spouses' taxes would be to replace the joint and several
liability standard with a proportionate liability standard. Under
such a standard, taxpayers would be responsible only for the taxes
generated by their individual incomes and assets or, for taxpayers
living in community property states, for the tax associated with
one-half of the community income. Options for administering a
proportionate liability standard include (1) eliminating joint
returns and requiring all taxpayers to file separately, (2) retaining
joint returns but modifying them so that each spouse's income and
deductions would be reported in separate columns, and (3) retaining
the current joint return requirements but applying proportionate
liability only in cases where there are unpaid taxes or subsequent
tax assessments. Each of these options represents a trade-off
between clearly establishing each taxpayer's liability and the amount
of paperwork and administrative burden created for taxpayers and IRS.
The separate and modified joint return options would increase
taxpayers' filing burden and IRS' return-processing costs. All three
options could increase the costs of IRS' audit, underreporter, and
collection programs.
Divorcing couples may specify in their divorce decrees how future
liabilities resulting from their prior joint returns are handled.
IRS officials at the local level said many taxpayers are surprised to
discover that IRS is not bound by these agreements because it is not
a party to the decree. Requiring IRS to be bound by divorce decrees
is impractical for two major reasons. First, federal tax matters are
the exclusive jurisdiction of certain federal courts, while divorce
matters are generally handled by state courts. Thus, there is
currently no legal forum where IRS and the parties to a divorce could
resolve issues relating to both tax matters and divorce proceedings.
Furthermore, this proposal could require IRS to become involved in
every divorce settlement or trial. In 1994, about 1.2 million
divorce decrees were granted in the United States. IRS officials
also raised related concerns, such as whether their interpretation of
lengthy and complex divorce decrees would increase the number of
appeals and whether divorce decrees would be manipulated to reduce
tax liabilities.
IRS can treat taxpayers living in community property states
differently from taxpayers living in common law states when
collecting taxes. Unlike in common law states, IRS can levy one
spouse's income to satisfy the premarital tax debts of the other
spouse in community property states because of the joint ownership of
property in those states. In contrast, IRS cannot levy the income of
one spouse to pay the premarital tax debts of the other spouse in
common law states because spouses do not have a legal entitlement to
each other's property. Since IRS does not maintain data on how often
these levy actions occur, we could not assess the potential impact on
IRS of changing the law to treat everyone the way it treats taxpayers
in common law states.
SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :3
To calculate the potential universe of innocent spouses, we used IRS'
Statistics of Income data for tax year 1992 to estimate the number of
joint returns filed that year and whether or not the taxpayers lived
in community property states. We used IRS' information on the
results of its 1992 underreporter program to estimate the number and
dollar amount of such assessments made against joint filers. We also
used the Audit Information Management System (AIMS) database to
identify the number and amount of audit assessments made against tax
year 1992 joint returns. Finally, we used data from the Bureau of
the Census and the Department of Health and Human Services to
calculate an annual divorce rate and estimate the number of joint
filers that divorce each year.
To determine IRS' practices and procedures for handling innocent
spouse cases, we interviewed IRS officials at headquarters, four
district offices, and three service centers to discuss their
procedures for identifying and processing innocent spouse cases. We
selected the Baltimore, Philadelphia, Arkansas-Oklahoma, and San
Francisco District Offices and the Philadelphia, Austin, and Fresno
Service Centers to give a diverse geographic perspective. We also
reviewed innocent spouse cases at the San Francisco District Office
and the Fresno Service Center located near our San Francisco office.
In addition, IRS, as part of its own efforts to assess the problems
related to divorced and separated taxpayers, had requested five
Problem Resolution Offices to forward the innocent spouse cases
handled between January and June 1996 to its Problem Resolution
Office in headquarters. We reviewed the 31 innocent spouse cases.
To determine whether the innocent spouse provisions provide the same
treatment to all taxpayers, we reviewed the literature examining the
provisions and IRS data, and compared the federal innocent spouse
provisions to state innocent spouse provisions.
To determine the potential effects of changing the current joint and
several liability standard to a proportionate liability standard, we
used IRS' Statistics of Income data for tax year 1992, underreporter
assessments,\5 and the AIMS database to estimate the number of
taxpayers who filed using the "married, filing jointly" status. We
also worked with IRS to develop an estimate of the number of
taxpayers who had an assessment made against a previously filed joint
return to estimate the universe of taxpayers who would have been
affected by any changes to the standard. In addition, we reviewed
proposed alternatives to the joint and several liability standard
prepared by the American Bar Association (ABA) and the American
Institute of Certified Public Accountants (AICPA). We also met with
North Dakota state officials, who administer a proportionate
liability standard on joint state income tax returns. To determine
the potential tax administration issues and taxpayer burden
associated with establishing a proportionate liability standard, we
developed a stratified probability sample of 200 joint tax returns
from IRS' tax year 1992 Statistics of Income file to estimate the
amount of income that could be identified as either joint or separate
income. We projected this sample to the universe of 48 million
taxpayers who used the "married, filing joint" status at a 95 percent
confidence level.
To determine the potential effects on IRS of requiring it to be bound
by divorce decrees, we analyzed the legal ramifications of binding
IRS to the terms of lower court decisions. We also discussed the
benefits and problems associated with following the provisions of
divorce decrees with officials from IRS and officials from
California, Wisconsin, and Delaware, whose state tax agencies are
bound by divorce decrees.
To determine the potential effects on IRS of changing the law to
limit its ability to attach community property, we discussed with IRS
officials the policies related to the attachment of income and assets
of one taxpayer to pay the debts incurred by his or her spouse before
the marriage.
We performed our review from February 1996 through September 1996 in
accordance with generally accepted government accounting standards.
We requested comments on a draft of this report from the Commissioner
of Internal Revenue, which we received on January 15, 1997. These
comments are discussed on pages 23 to 25, and a copy of the comments
are included in appendix VI.
--------------------
\5 In its underreporter program, IRS computers match income shown on
information returns (e.g., forms W-2 and 1099 for interest and
dividends) with income that taxpayers report on their tax returns to
determine whether taxpayers reported all their income. When
discrepancies are found, IRS contacts taxpayers to resolve the issue
and assess additional taxes, if required.
ESTIMATED UNIVERSE OF POTENTIAL
INNOCENT SPOUSES
------------------------------------------------------------ Letter :4
Because IRS did not have data on the number of innocent spouse
requests filed, we developed an estimate of the potential universe of
taxpayers that could qualify under the current innocent spouse
provisions. We estimated that a spouse from up to 587,000 couples
may have been eligible for innocent spouse relief in 1992.
About 48 million joint tax returns were filed for tax year 1992.
From IRS' data on tax year 1992 audit and underreporter programs, we
estimated that 1.25 million couples filing joint returns were
assessed additional taxes under these programs--250,000 were audit
assessments and 1 million were underreporter assessments. Of these
1.25 million returns, about 587,000 had additional tax assessments
exceeding $500, which is the minimum dollar threshold required for
innocent spouse relief. Appendix I describes the methodology we used
to make these estimates.
However, our estimate of 587,000 couples represents the maximum
number of taxpayers potentially eligible for innocent spouse relief;
fewer would probably actually qualify. Some couples would also
probably have been assessed additional taxes as a result of
overstated deductions, credits, or basis, which have other dollar
thresholds in addition to the $500 threshold. Data were not
available that we would need to estimate the number of joint
taxpayers whose tax year 1992 additional tax assessments resulted
from overstated deductions, credits, or basis. Also, some of the
587,000 couples may not have qualified for innocent spouse protection
because they knew there was a substantial tax understatement. This
knowledge would have made them ineligible for relief even if the tax
deficiency was solely attributable to the actions of one spouse.
Although an unknown number of the 587,000 couples could potentially
seek innocent spouse relief, IRS officials told us that the severity
of the problem for taxpayers seeking such relief is much greater in
the case of divorced or separated taxpayers. Therefore, we also
estimated the number of taxpayers who could potentially be eligible
for relief and may have divorced during the 3 years since the 1992
joint returns were filed. Using a 2 percent per year divorce rate,
we estimated that 35,000 divorced taxpayers might have been eligible
for innocent spouse relief for additional tax assessments of more
than $500. Appendix I describes how we developed this estimate.
LIMITED DATA ON INNOCENT SPOUSE
CLAIMS WERE AVAILABLE
------------------------------------------------------------ Letter :5
IRS did not accumulate data on the number of cases requesting
innocent spouse relief or the number of cases for which it grants
such relief. Although IRS did not systematically collect data on
innocent spouse cases, we found that some IRS operating units we
visited maintained some information on the innocent spouse cases they
handled. The limited information found on innocent spouse claims
indicated that few requests were received, and of these, most were
denied. For example, records at the Fresno Service Center indicate
it received 90 Offer in Compromise cases requesting innocent spouse
relief during the 3 years between March 1993 and February 1996.\6 The
service center denied 48 of these requests because either (1) they
dealt with taxes reported as owed on the original return but not paid
rather than with subsequent assessments; or (2) the issues causing
the assessment had already been resolved by IRS' Appeals Division or
the Tax Court, and the claimant had agreed to the decisions. The
remaining 42 cases were referred to district offices for processing.
As of September 1996, 26 of these 42 cases had been resolved, with 7
being allowed and 19 being denied.
We found that most of the cases handled by IRS' Problem Resolution
Office\7 were also denied. In fiscal year 1996, the Problem
Resolution Office requested information from five offices on innocent
spouse cases. We reviewed 31 Problem Resolution Office cases from 4
district offices and 1 service center where taxpayers raised the
innocent spouse issue between January and June 1996. For the 21
cases where a decision had been reached, IRS granted relief in 10
cases. Appendix II shows our analysis of the Problem Resolution
cases we reviewed and summaries of some of the cases.
Furthermore, according to IRS officials, the Tax Court denied relief
in one-third of the innocent spouse cases decided in 1996.
--------------------
\6 IRS identified the 90 Offer in Compromise cases by a manual review
of Offer in Compromise correspondence files maintained in the Joint
Compliance Branch. Although other requests for innocent spouse
relief may have been received at the Fresno Service Center during
this time period, we were unable to identify any other operating
units at the center that maintained such information.
\7 The Problem Resolution Program, administered by IRS' Problem
Resolution Office, is designed to help taxpayers who have been unable
to resolve their tax problems after repeated attempts to do so with
another IRS unit.
INFORMATION AVAILABLE ON
APPLYING FOR INNOCENT SPOUSE
RELIEF WAS LIMITED
------------------------------------------------------------ Letter :6
Although innocent spouse relief is clearly established in law and
regulation, we observed that little information about the criteria
for granting it or how to apply for it was available from IRS
sources. The innocent spouse relief provisions are described in
several IRS publications, including Your Federal Income Tax
(Publication 17), Divorced or Separated Individuals (Publication
504), and Exemptions, Standard Deduction, and Filing Information
(Publication 501). However, these publications do not provide any
guidance on how to request relief. Furthermore, they are developed
to help taxpayers prepare their returns, which is far in advance of
the time that taxpayers might need information on the possibility of
innocent spouse relief.\8 In contrast, Examination of Returns, Appeal
Rights, and Claims for Refund (Publication 556) and Understanding The
Collection Process (Publication 594) are totally silent about
innocent spouse relief, even though these publications are more
directly related to the procedures that apply when taxpayers are
billed for their spouses' taxes.
--------------------
\8 IRS generally does not begin auditing tax returns until about 6
months after the April 15 return due date. Also, it generally does
not begin investigating cases in its underreporter program until 7
months after the April 15 filing date.
IRS HAD NO WELL-DEFINED
PROCEDURES FOR REQUESTING
AND PROCESSING INNOCENT
SPOUSE CLAIMS
---------------------------------------------------------- Letter :6.1
IRS also lacks well-defined procedures for taxpayers to request
innocent spouse relief. As noted, in those limited cases where IRS
has publicized the innocent spouse provisions, there is no guidance
as to how taxpayers should request such relief. In those innocent
spouse cases we were able to identify and review, we found no
consistent process for claiming relief. In most cases, we found that
either the taxpayers or their representatives had (1) contacted
Problem Resolution Offices, which were established to assist
taxpayers who cannot resolve their problems through normal IRS
channels; or (2) requested relief through an Offer in Compromise,
which is used in the cases of taxpayers who cannot pay the full
amount of the balance due and decide to offer a lesser amount. The
fact that taxpayers are commonly using these two approaches to seek
innocent spouse relief may indicate that taxpayers are not provided
with adequate guidance for seeking relief by IRS.
In contrast, we found a much more taxpayer-friendly approach taken in
the case of "injured spouse" claims. Injured spouses are joint
filers whose joint refunds have been seized to pay certain of their
spouses' nontax debts, such as past-due child or spousal support or a
federal loan. Injured spouses can apply for their portion of the
joint refund by completing Form 8379, Injured Spouse Claim and
Allocation. The 1040 instruction booklet provides a clear
explanation of the injured spouse provisions, the qualifying
conditions to be met, the specific form to be prepared, and how the
claim should be filed with IRS. The information is provided under
the refund line item instructions and tells taxpayers to attach the
Form 8379 to their tax return.
We also found confusion within IRS about how taxpayers should request
innocent spouse relief. The various IRS units we contacted to
determine the procedures they followed in handling innocent spouse
cases took different approaches to considering relief. For example,
two district offices granted relief using Offers in Compromise based
on doubt of liability, while staff at one service center routinely
denied such requests as inappropriate. This latter service center
staff's position was that Offers in Compromise based on doubt of
liability can be used only to argue that IRS miscalculated the tax
assessment, not to argue that a taxpayer is not liable for paying the
assessment. An official at one district office said he would advise
taxpayers to fill out an injured spouse form, while an official at a
service center said he was unaware that innocent spouse relief
existed.
MODIFYING TAX CODE PROVISIONS
COULD ALLOW MORE TAXPAYERS TO
QUALIFY FOR RELIEF
------------------------------------------------------------ Letter :7
Critics of the innocent spouse provisions, such as ABA and AICPA,
contend that the current provisions do not ensure that taxpayers
receive equitable relief. For example, they said that the dollar
thresholds included in the provisions result in eligibility criteria
for relief based on income and not strictly on whether the spouse was
innocent. Also, under the current provisions, spouses can receive
relief if deductions have absolutely no basis in fact or law, but not
if the deductions are simply erroneous. Furthermore, spouses
requesting relief must establish that they did not know and had no
reason to know their spouses were cheating on the taxes. Critics
note that the concept of "no reason to know" has not only been
interpreted differently by various courts, but is difficult for the
potential innocent spouse to prove. Appendix II includes several
examples that illustrate the issues involved in applying the innocent
spouse provisions.
According to IRS officials, Congress required innocent spouses to
meet certain dollar thresholds to qualify for relief as a way of
filtering out insignificant claims. The dollar thresholds clearly
exclude some taxpayers from relief and may be inconsistent with the
goal of providing relief to deserving taxpayers. In the case of a
tax assessment related to an improper deduction, credit, or basis,
the amount of the assessment must exceed 10 percent of the claimant's
adjusted gross income (AGI) for the claimant's most recent tax year
if such income is $20,000 or less, but the assessment must exceed 25
percent of AGI if it is more than $20,000. Thus, if a taxpayer's AGI
is $20,000 or less, relief could be available on assessments of
$2,000 or less, but if the taxpayer's AGI is more than $20,000,
relief would be available only if the assessment exceeded $5,000.
This distinction appears to be more related to an ability to pay or
degree of hardship than to the innocence of the taxpayer. The logic
behind these dollar thresholds is clouded even more because the
potential innocent spouse's AGI is based on the tax year ending
before the notice of deficiency (which may be several years after the
tax year of the joint return) and must include the income of any new
spouse whether or not a joint return was filed. Finally, the dollar
thresholds prevent taxpayers with smaller liabilities from obtaining
relief, since the minimum understatement of tax in all cases must be
more than $500, which could preclude lower income taxpayers from
obtaining relief.
The 1984 Tax Reform Act extended relief to include deductions but
requires that such items be grossly erroneous, meaning there is no
basis in fact or law for the claim. The distinction between a
deduction having no basis in law or fact versus its just being
erroneous is difficult to comprehend and can lead to various
interpretations by IRS and the courts. This problem is compounded by
the fact that IRS' regulations governing innocent spouse relief were
issued in 1974 and have not been updated since that time to
incorporate the changes to the innocent spouse provisions resulting
from the 1984 Tax Reform Act.
The "knowledge" factor is perhaps the most subjective element in the
current innocent spouse provisions. For someone to prove that they
did not know and had no reason to know of a financial transaction
undertaken by his or her spouse would generally be difficult, if not
impossible. IRS and the courts consider circumstantial factors, such
as education, involvement in the family's financial affairs, and
lifestyle, in assessing this contention. For example, one indicator
that IRS uses to determine if spouses were aware of the tax avoidance
is whether they benefited by living a lifestyle significantly better
than could be supported by the reported income. However, according
to critics, a determination of whether a taxpayer's lifestyle was
significantly better because of the tax avoidance is fairly
subjective and the courts have interpreted the criteria differently.
Some district offices have designated an employee as the innocent
spouse coordinator so that only one employee applies the knowledge
test.
Some states have decided not to apply the federal innocent spouse
provisions and have modified them for state tax purposes. Our survey
of the District of Columbia and the 43 states that have personal
income taxes showed that 20 do not have innocent spouse provisions,
18 have innocent spouse provisions based on the Internal Revenue
Code, and 6 have less restrictive provisions (see app. III for a
listing of the states).
California, Iowa, Louisiana, and Oklahoma do not apply dollar
thresholds when granting innocent spouse relief, while New York
applies a $100 threshold. Massachusetts does not apply the
percentage of income threshold for taxes resulting from grossly
erroneous deductions. In addition, California and Oklahoma do not
require that deductions have no basis in fact or law, simply that
they be erroneous.
We estimated that if the federal innocent spouse tax code provisions
had been modified to eliminate the dollar thresholds as was done by
California, Iowa, Louisiana, and Oklahoma, the maximum number of
couples filing tax year 1992 returns potentially eligible for
innocent spouse relief would have been 710,000. Assuming a divorce
rate of 2 percent per year, we estimate that 42,600 of these couples
would have divorced within 3 years.
Modifying the provisions to allow more taxpayers to qualify for
innocent spouse relief would likely result in some revenue loss
because IRS might not always be successful in collecting from the
culpable spouse. However, since IRS does not maintain data on how
often it collects from the culpable spouse, we could not estimate the
size of the potential revenue loss. IRS would also incur some
additional collection costs associated with pursuing the culpable
spouse.
POTENTIAL IMPACT OF REPLACING
THE JOINT AND SEVERAL LIABILITY
STANDARD WITH PROPORTIONATE
LIABILITY
------------------------------------------------------------ Letter :8
An alternative way to ensure that taxpayers are not held liable for
their spouses' taxes would be to replace the joint and several
liability standard with a proportionate liability standard. Under
the generally accepted definition of proportionate liability,
taxpayers would be held responsible only for the taxes generated by
their own individual incomes and assets or, for taxpayers living in
community property states, for the tax associated with one-half of
the community income. We identified three options for administering
a proportionate liability standard that represent trade-offs between
clearly establishing each taxpayer's liability on their tax returns
and the amount of paperwork and administrative burden created for
taxpayers and IRS. Two options in addition to proportionate
liability that would limit IRS' ability to hold taxpayers liable for
their spouses' taxes are to (1) allow taxpayers to amend their filing
status after the due date of the return and (2) have each taxpayer
identify on the return the percentage of the total liability for
which he or she would be responsible.
COMPARISON OF THE POTENTIAL
IMPACT OF THREE
ADMINISTRATIVE OPTIONS ON
TAXPAYERS AND IRS
---------------------------------------------------------- Letter :8.1
Our review of the literature identified three options for
administering a proportionate liability standard. The options are to
(1) eliminate joint returns and require all taxpayers to file
separately,\9 (2) retain joint returns but modify them so that each
spouse's income and deductions are reported in separate columns, and
(3) retain the current joint return requirements but apply
proportionate liability only in cases where there are delinquent
taxes or subsequent tax assessments. We evaluated the potential
effects of these options on IRS' tax administration processes and
taxpayers' burden. We did not consider how these options would
potentially affect tax revenues. Table 1 shows the pros and cons of
the three options for taxpayers and IRS.
Table 1
Pros and Cons of Different Methods of
Administering a Proportionate Liability
Standard
Separate Modified joint Current joint
return return return
Entity option\a option\b option\c
---------------------- -------------- -------------- --------------
Taxpayers
Pros If divorced, If divorced, No additional
individual individual paperwork
liability is liability is burden.
more clearly more clearly
Cons established. established.
Must establish
Must prepare Must allocate individual
two returns joint income, liability if
but receive deductions, additional
limited or no and credits taxes
benefit while but receive assessed.
married. limited or no
benefit while
married.
May have a
higher tax May have a
liability higher tax
under current liability
rate under current
structure. rate
structure.
IRS
Pros Individual Individual No additional
liability more liability more return-
clearly clearly processing
established. established. costs.
Cons
Increased Might increase Must establish
costs for costs for individual
processing up keying liability if
to twice as additional additional
many returns data into taxes
for married computer assessed.
couples. systems.
Increased Increased
difficulty in difficulty in
matching matching
income income
reported on reported on Increased
returns to returns to collection
information information costs because
returns. returns. IRS would have
to collect
Increased Increased from each
collection collection taxpayer
costs because costs because rather than a
IRS would have IRS would have couple.
to collect to collect
from each from each
taxpayer taxpayer
rather than a rather than a
couple. couple.
----------------------------------------------------------------------
\a Each spouse files separate return.
\b Income split out separately on joint return.
\c Proportionate income only for returns with unpaid taxes or
subsequent tax assessments.
Source: GAO's analysis of three proportionate liability options.
As shown in table 1, establishing proportionate liability on either a
separate or joint return is of limited or no benefit to married
taxpayers who generally would pay their income tax from joint assets
or, if they prefer, already have the option of choosing the "married,
filing separately" option to limit their liability to their own
income. However, such a system would clearly establish liability so
that in the event of a tax shortfall, divorced taxpayers could more
easily establish the extent to which they are liable for the unpaid
taxes or assessments. This clarity, however, is purchased at the
cost of a significant burden for taxpayers. For example, in tax year
1992, about 48 million couples filed joint returns but only about 4.5
million, or 9 percent, of them had unpaid tax liabilities or
subsequent tax assessments. Thus, under the separate and modified
joint return options, about 43 million more couples would have been
burdened by proportionate liability than would have potentially
benefited. These taxpayers would, at a minimum, have had to allocate
their income, deductions, and credits on a joint return and, if
required to file separately, file an additional return. Under the
current joint return option, however, only the 4.5 million couples
with unpaid tax liabilities or additional tax assessments would
potentially have had to proportion their income, deductions, and
credits. Since, according to IRS, married couples are less likely to
request a proportionate split of their joint tax liability even if
one of the spouses is innocent, this number may overestimate how many
taxpayers would have actually benefited. We estimate that about
270,000 taxpayers would divorce during the 3 years after the returns
were filed.
Furthermore, as we reported in September 1996,\10 most married
taxpayers would pay higher total taxes if they filed separately
rather than jointly. As a result, requiring these taxpayers to file
separately could create a higher tax liability for a significant
number of taxpayers under the current rate structure.
IRS would also face increased return-processing costs under the
separate return or modified joint return-filing options, but not
under the current joint return option. For example, if taxpayers
were required to file separately, IRS would have to process up to 48
million additional returns since each dual-income couple who formerly
filed a single joint return would have to file two returns. We
estimate that if all 48 million joint filers had to file two returns,
it would cost IRS an additional $199 million to process the
additional tax returns. If married taxpayers were allowed to file
jointly but had to report their income and deductions separately on
the return--for example, a tax return column for each spouse--IRS
might have to make twice as many data transcription entries as it
currently does. If all 48 million joint filers reported two income
streams, we estimate that the additional data entry costs could be
about $19 million.\11 Appendix IV describes our methodology for
estimating these costs. IRS' underreporter program might also face
additional computer matches, but we did not try to estimate these
increased costs. There would be no additional return-processing
costs under the current joint return option.
IRS' tax compliance costs would increase under all three options;
however, it would experience significantly more costs under the
separate and modified joint return options than under the current
joint return option. For example, IRS' underreporter program costs
could increase because it might have trouble matching proportionate
tax returns (i.e., separate returns and modified joint returns) to
information returns. In our review of a stratified probability
sample of 200 joint tax returns for 1992 and the information returns
associated with these returns, we found that 77 percent of the income
reported on the returns was separately held and IRS would have little
difficulty allocating this income. About 2 percent of the income was
reported as joint income, and we could not determine whether 12
percent of the income was jointly or separately held. As a result,
IRS would generally have difficulty allocating this income. (App. V
lists the various income types and whether they were jointly or
separately held.)
For separate returns, IRS would not be able to readily match jointly
held income to determine whether all the income was reported unless
it cross-referenced the return to the spouse's return. IRS would not
have this problem with the modified joint return because both
spouses' income on the joint return could be totaled by computer and
matched to information returns. However, under both the separate and
modified joint return options, IRS could not easily determine whether
jointly held income was correctly apportioned without requiring
taxpayers to either document the information provided on the return
or to maintain separate information return accounts for all their
income.
The advantage of establishing individual tax liability only if there
was a tax shortfall is that it would create little additional burden
for taxpayers or IRS. The disadvantage is that in the event of a tax
shortfall, the taxpayers and IRS would have to apportion a jointly
reported tax liability between the two taxpayers who signed the
return. Since the current information reporting system shows certain
income and deductions jointly, Congress or IRS would need to develop
guidelines on how such income and deductions should be handled.
In addition to increased underreporter costs, IRS would likely face
increased collection costs because it would have to collect from each
taxpayer rather than the couple or whichever taxpayer it found first.
--------------------
\9 Married couples currently have the option of filing separately,
which in effect proportions their tax liability to their own income,
unless the couple lives in a community property state, in which case
each taxpayer is liable for the taxes associated with any separate
income they may have and one-half of the community income. However,
only 5 percent of married taxpayers file separately, because the tax
rates are generally lower if they file a joint return.
\10 Tax Administration: Income Tax Treatment of Married and Single
Individuals (GAO/GGD-96-175, Sept. 3, 1996).
\11 If the modified return had separate columns for each spouse and a
total column, IRS might need to key in only the total column, with
the other data being used only if a claim for innocent spouse relief
arises.
PROPORTIONATE LIABILITY
ESTABLISHED ONLY FOR UNPAID
TAXES OR AUDIT ASSESSMENTS
WOULD BE LEAST COSTLY OPTION
---------------------------------------------------------- Letter :8.2
The least costly and disruptive of the three proportionate liability
options would be to retain the current joint return and proportion
income and deductions only in cases where there are either unpaid
taxes or additional tax assessments. This option is endorsed by ABA
and is practiced by North Dakota--the only state that has a
proportionate liability standard. Each of these entities, however,
advocates different methods for proportioning tax liabilities.
ABA, which advocates a proportionate liability standard, proposes
apportioning the liability reported on the original return by
calculating (1) the spouses' taxes as if they had filed using the
"married, filing separately" status; and (2) the percentage of the
couple's combined taxes at the separate rate attributable to each
spouse. IRS would then apply each spouse's percentage to the joint
tax to determine each taxpayer's liability. The burden of proof for
calculating the proportionate liability would fall on the taxpayer.
If IRS assesses additional taxes through an audit, the deficiency
would be the liability of whichever spouse generated the unreported
income or disallowed deduction, thus reducing the need for innocent
spouse relief. IRS officials believe that innocent spouse relief
would still be necessary in certain circumstances, such as when one
spouse purchases rental property in both spouses' names and does not
inform the other spouse of the asset or income.
North Dakota is the only state that applies a proportionate liability
standard to the state income tax return. However, in practice, the
North Dakota Tax Commission administers the state tax system as if a
joint and several liability standard applied to the state joint
returns because it pursues whichever taxpayer it finds first to
collect the tax liability. If a divorced taxpayer claims that the
tax liability is really attributable to the ex-spouse, the state
applies a proportionate liability standard to the joint return. The
state uses information returns from IRS to initially apportion
reported income and deductions. For income reported in both
taxpayers' names, the state assesses 100 percent of the joint income
to both taxpayers, and the burden then falls on the taxpayer to
document a different allocation. Joint deductions, such as those for
dependents, are allocated in proportion to the amount of income each
spouse generated. In other words, if a spouse generated 40 percent
of the income, that spouse could claim 40 percent of the deduction.
Unreported income is attributed to whoever earned the income.
North Dakota Tax Commission officials said that although
proportionate liability addresses the problems of some divorced
taxpayers, it is not a panacea that produces a just result in all
cases. For example, state officials noted that most farm land is
held jointly, but the income is claimed only by the farming husband.
They questioned the fairness of allowing a wife to disclaim any
responsibility for income generated by an asset that was owned by
both spouses. State officials said they were also bothered that
nonworking taxpayers who are completely supported by their spouses
can then disclaim any responsibility for their spouse's taxes. Such
disclaimers are particularly ironic in that nonworking spouses
generally establish their claims to assets during divorce proceedings
by arguing that they enabled their spouses to generate income. State
officials questioned the fairness of allowing taxpayers to make such
arguments to bolster their claims to assets and then to escape any
tax liability by arguing the income was not theirs.
OTHER ALTERNATIVES TO
PROPORTIONATE INCOME
---------------------------------------------------------- Letter :8.3
Two other alternatives to proportionate liability have been proposed
or adopted at the state level that would limit IRS' ability to hold
taxpayers liable for their spouses' taxes. First, IRS could allow
taxpayers to amend their filing status after the due date of the
return. Nine states give taxpayers the opportunity to do this.
Second, AICPA advocates replacing the joint and several liability
standard with an allocated liability standard. Under this standard,
taxpayers would specify on the return a percentage of the total
liability for which they would each be responsible. If the taxpayers
failed to specify an allocation, they would each be responsible for
one-half of the tax liability.
BINDING IRS TO DIVORCE DECREES
WOULD BE IMPRACTICAL
------------------------------------------------------------ Letter :9
Divorcing couples may specify in their divorce decrees how future
liabilities resulting from their prior joint returns are handled,
i.e., one spouse is entirely liable, both spouses are equally liable,
or some other permutation. However, IRS is not bound by these
divorce decrees because it is not a party to the decree. IRS
officials at the local level said many taxpayers are surprised to
discover that IRS is not bound by divorce decrees. According to ABA
representatives, many divorce attorneys are not well-versed in tax
matters and do not realize that divorce decrees do not provide
adequate protection against additional federal taxes. According to
IRS officials and private sector practitioners, IRS' practices in
this regard are similar to the practices of private sector creditors,
such as credit card companies, which do not feel they are bound by
divorce decrees when collecting on joint debts.
According to certain members of ABA's Committee on Domestic
Relations, a legislative change to bind IRS to divorce decrees
appears to be impractical for two major reasons. First, current
federal law provides no mechanism whereby IRS can be a party to
divorce proceedings. Federal tax matters are the exclusive
jurisdiction of the U.S. Tax Court, federal district courts, the
U.S. Claims Court, and the U.S. Bankruptcy Court. Divorce matters,
however, are generally handled by state courts. Federal courts have
traditionally refused to consider any legal action involving divorce.
Thus, providing a legal forum where IRS and the parties to a divorce
could resolve issues relating to both tax matters and divorce
proceedings would require a fundamental and extensive change in
either federal tax law or state domestic relations law.\12
Second, binding IRS to divorce decrees could require IRS to become
involved in every divorce settlement or trial. In 1994, about 1.2
million divorce decrees were granted in the United States. To be a
party to this many legal proceedings nationwide each year would
create a significant administrative burden for IRS. For example, the
California Franchise Tax Board is bound by divorce decrees if the
decree (1) does not reduce liability for income under the exclusive
management and control of the spouse, (2) does not affect taxes that
have already been paid, and (3) has been cleared by the California
Franchise Tax Board when reportable gross income exceeds $50,000 or
the tax liability exceeds $2,500. Once the California Franchise Tax
Board has approved the proposed tax allocation, the state becomes a
party to the decree and is bound by it. State officials said the
requirement has increased demands on administrative resources, and
they had a backlog of about 100 requests for approval of tax
allocations.
The Delaware Division of Revenue is also bound by divorce decrees,
even though it is not a party to the decree. As a state agency, the
Delaware Division of Revenue considers itself bound by state court
decisions. However, most Delaware taxpayers file separately because
of the state's rate structure, so the requirement creates little
administrative burden. The Wisconsin legislature recently passed
legislation requiring the Department of Revenue to be bound by
divorce decrees beginning in the next tax year. Wisconsin officials
are just starting to consider the consequences of this legislation on
processing returns and revenue collection.
IRS officials also believe the number of appeals would increase
because divorce decrees can be lengthy and complex documents that are
open to more than one interpretation. Furthermore, IRS officials
fear that divorce decrees would be manipulated to reduce tax
liabilities. For example, one spouse might retain sole ownership of
the couple's residence, the couple's major asset, while the spouse
without assets takes responsibility for the taxes. Thus, IRS would
not be able to place a lien against the residence to force collection
action for any delinquent taxes.
--------------------
\12 The views cited in this paragraph were submitted to the IRS in
comments by ABA's Committee on Domestic Relations in response to
Notice 96-19. The comments specifically noted that the positions
taken represented the individual views of the members who prepared it
and did not represent the position of ABA or of the Section on
Taxation. With respect to the specific impact of such a legislative
change on taxpayers or IRS in each state, the comments also noted
that to answer the question definitively would require a law survey
of all 50 states, which the Committee had not at that time done. We
also have not done a law survey of the 50 states and, therefore, also
cannot answer this question definitively.
IRS FOLLOWS STATE PROPERTY LAWS
IN COLLECTING PREMARITAL TAX
DEBTS
----------------------------------------------------------- Letter :10
About 13 million, or 27 percent, of all taxpayers who filed joint
returns in 1992 lived in community property states. Some of these
taxpayers may have been faced with tax liabilities incurred by their
spouses before their marriage, which they would not have encountered
if they lived in a common law state. This disparate treatment
between taxpayers residing in community property states versus those
living in common law states occurs because IRS, as with other
creditors, follows state law in classifying married couples' rights
in property.
Because the income, including wages, of taxpayers living in certain
community property states is considered community property, IRS can
place a levy on the wages or other separate income of either spouse
to satisfy an existing tax debt, even if that tax debt was incurred
by the other spouse before their marriage. In contrast, IRS cannot
place a levy on the separate income of one spouse to pay the taxes
due from the other spouse in a common law state. Once the income of
either spouse is placed in a joint account it would be subject to IRS
seizure in both community property and common law states. However,
the placement of wages into a joint account is a matter of choice on
the part of the taxpayers and can be avoided if they so choose.
According to IRS officials, the agency does not have specific
procedures for placing levies on a spouse's income for premarital
taxes incurred by the other spouse. Officials told us that because
community property laws differ from state to state, it is up to IRS
collections personnel in each community property state to determine
whether they will levy a spouse's income. Nonetheless, under IRS'
collection procedures, levy action is generally to be taken against
the individually held income, such as wages, of the taxpayer who
incurred the tax debt or any jointly held income, such as an
interest-bearing account, not against the separate income of the
other spouse. In 1994, IRS issued 3 million third-party levies.
However, the agency did not know how many of these levies were placed
on the income of a spouse living in a community property state to pay
the premarital tax debts of the other spouse. As a result, we could
not assess the effects of changing the law to prevent IRS from making
these types of levies.
The current tax code contains several provisions that override state
community property laws for noncollection activities. For example,
the innocent spouse provisions, section 6013(e), specifically prevent
IRS from following state community property laws in determining the
tax liability on omission from gross income. Also, under section
879(a), community property laws do not apply to the earned income of
nonresident aliens. However, making an exemption for this specific
collection activity for premarital debts would set a precedent
because it would require IRS to ignore state community property laws
for a collection activity. Furthermore, unless the states change
their community property laws, such a change would not necessarily
protect taxpayers' assets. Private sector creditors could continue
to pursue community assets to satisfy community debts even though IRS
was precluded from attaching such assets.
The inequities created by having IRS follow state community property
laws are not limited to levying the income of a spouse to pay the
premarital tax debts of the other spouse. For example, the
eligibility criteria for injured spouse relief on IRS Form 8379,
Injured Spouse Claim and Allocation, specifically state that
"Overpayments involving community property states will be allocated
by the IRS according to state law. Claims from California, Idaho,
Louisiana, and Texas will usually result in no refund for the injured
spouse." Thus, IRS follows state community property laws when
withholding refunds to apply against nontax debts.
CONCLUSIONS
----------------------------------------------------------- Letter :11
IRS does not track how often innocent spouse relief is requested or
granted; however, we estimate that a relatively small number of
taxpayers are eligible for relief under the current innocent spouse
provisions. The provisions, with their various qualifying criteria,
are complex and difficult to meet, and they require IRS staff to
weigh various factors in deciding whether to enforce the joint and
several liability standard or to relieve taxpayers of their joint
liability. IRS provides limited information to taxpayers about the
availability of and criteria for relief and does not have a tax form
and procedures for requesting and granting relief.
Several proposals have been made to modify innocent spouse relief.
More taxpayers might qualify for relief if Congress eliminated or
modified the dollar thresholds to allow IRS to consider relief for
taxpayers with liabilities less than $500. Replacing the joint and
several liability standard with a proportionate liability standard
may also provide additional relief to taxpayers, but this alternative
could create a significant administrative burden for taxpayers and
IRS. Requiring IRS to be bound by divorce decrees appears to be
impractical because of the legal and administrative hurdles that
would have to be resolved. Finally, if the law were changed to
prevent IRS from levying the income of one spouse to pay the
premarital tax debts of the other spouse, the inequities between
taxpayers in community property states and common law states might be
reduced. However, no data were available to assess the impact of
such a change.
RECOMMENDATIONS
----------------------------------------------------------- Letter :12
To improve IRS' administration of the innocent spouse provisions of
the tax code, we recommend that the Commissioner of Internal Revenue
(1) develop new or modify existing publications to better inform and
educate taxpayers about the availability of and criteria for innocent
spouse relief and, as part of that effort; (2) develop a tax form and
procedures for requesting and either granting or denying innocent
spouse relief. We also recommend that the Commissioner provide
additional internal guidance to IRS employees so that greater
consistency in processing innocent spouse cases can be achieved,
establish a cost-effective process for monitoring the consistency
being achieved, and update IRS' regulations to reflect current
requirements.
AGENCY COMMENTS AND OUR
EVALUATION
----------------------------------------------------------- Letter :13
In written comments on a draft of this report (see app. VI), the
Commissioner of Internal Revenue generally agreed with our
conclusions and recommendations and said that the findings in our
report present a real possibility for imminent legislative reform of
the innocent spouse provisions. She said that, thus, it would be
preferable to determine if the current legislative session of
Congress produces such reforms before devoting the resources
necessary to carry out our recommendations.
With this overall caveat, the Commissioner agreed with our first
recommendation that IRS should modify existing publications to better
inform taxpayers about the availability of innocent spouse relief.
Likewise, the Commissioner agreed with our second recommendation that
IRS develop a tax form and procedures for taxpayers to request
innocent spouse relief and for IRS to either grant or deny such
relief. She pointed out, however, that implementation of this
recommendation would require approval and clearance of such a form
from the Office of Management and Budget.
The Commissioner also agreed with our third recommendation that IRS
provide additional internal guidance to employees so that greater
consistency in processing innocent spouse cases can be achieved. She
said that the form and accompanying procedures developed under our
second recommendation would also largely meet the intent of the third
recommendation.
The Commissioner said that she understood the concern that prompted
our fourth recommendation to establish a cost-effective process for
monitoring consistency in processing innocent spouse cases. She
said, however, that existing management and information systems do
not have the capability to track issues on specific cases and that
the only alternative way to accomplish the recommendation, manual
tracking of such cases, would be cost prohibitive and less reliable.
It appears that the Commissioner misinterpreted the intent of our
recommendation. We fully agree with her points about the capability
of existing systems and the cost of manual tracking for such cases,
and in fact, we considered these issues in arriving at the wording of
our recommendation. Recognizing the problems with its existing
systems and manual tracking, our intent was to have IRS explore
process options that would not be cost prohibitive but would
facilitate a greater level of consistency in innocent spouse
decisions. We thought that such exploration was needed because we
found a substantial amount of inconsistency and subjectivity, both
within and among IRS districts, on how to interpret the knowledge
factor in the current innocent spouse provisions.
As an example of the kind of option that we thought IRS might explore
in response to our recommendation, we noted that some districts had
designated an employee as the innocent spouse coordinator, so that
only one employee in each of those districts applied the knowledge
test. This caused us to think that IRS could explore this as an
option by first determining whether the districts with such
coordinators have achieved greater consistency than those without
coordinators. If this proves to be so, IRS could then explore the
cost effectiveness of establishing innocent spouse coordinators in
the remaining districts. By serving as focal points and monitors in
their respective districts, these coordinators might facilitate more
consistent decisions within districts, and by networking with each
other, they might facilitate consistency among the districts. There
may well be other options that IRS could explore before reaching a
final decision on the best approach and it was to that end that we
worded our recommendation to provide IRS with latitude to decide how
it goes about achieving a greater level of consistency in its
administration of the innocent spouse provisions.
Finally, the Commissioner agreed in principle that IRS should update
its innocent spouse regulations to reflect current law, but said that
such action would be premature until Congress determines whether to
reform the existing provisions. We agree with the Commissioner's
position on this recommendation as well as with her overall statement
that it makes sense to wait until Congress decides whether to modify
the existing law in the current legislative session before devoting
the resources necessary to carry out most of our recommendations.
--------------------------------------------------------- Letter :13.1
We are sending copies of this report to various other congressional
committees, the Secretary of the Treasury, the Commissioner of
Internal Revenue, and other interested parties. Copies will be made
available to others upon request.
Please contact me at (202) 512-8633 if you or your staff have any
questions. Major contributors to this report are listed in appendix
VII.
Lynda D. Willis
Director, Tax Policy
and Administration Issues
ESTIMATED NUMBER OF POTENTIAL
INNOCENT SPOUSE CASES
=========================================================== Appendix I
This appendix describes how we derived the potential universe of
taxpayers that could be covered by the current innocent spouse
provisions in the Internal Revenue Code. The innocent spouse
provisions cover certain taxpayers who are assessed additional taxes
of more than $500 after they file a "married, filing joint" tax
return. Additional tax assessments generally result from an IRS
audit or from underreported income detected in IRS' underreporter
program. We used tax year 1992 audit and underreporter results to
estimate the potential innocent spouse universe because that was the
latest tax year for which most tax assessments had been completed.
ESTIMATED NUMBER OF INNOCENT
SPOUSE CASES THAT COULD RESULT
FROM AN AUDIT
--------------------------------------------------------- Appendix I:1
To determine the number of married couples filing joint returns that
had subsequent audit assessments, we used IRS' Audit Information
Management System (AIMS) database of tax year 1992 audit cases that
were closed during fiscal years 1993, 1994, 1995, and the first 9
months of fiscal year 1996. The AIMS data showed that for tax year
1992, 441,224 "married, filing joint" returns were audited and closed
by IRS' Examination Division. However, 190,393 of the 441,224 audits
resulted in either no change to the tax liability or in a refund.
The remaining 250,831 audits resulted in additional tax assessments.
Table I.1 shows the number of tax year 1992 "married, filing joint"
return audit cases with additional tax assessments that were closed
each fiscal year and shows whether the assessments were for amounts
of $500 and less or for more than $500.
Table I.1
Tax Year 1992 Closed Audit Cases for
"Married, Filing Joint" Returns
Returns with tax assessments
-------------------------------------------
More than
Fiscal year closed $500 or less $500 Total
------------------------- ------------- ------------- =============
1993 1,277 2,604 3,881
1994 8,876 40,407 49,283
1995 17,426 138,395 155,821
1996 4,785 37,061 41,846
======================================================================
Total 32,364 218,467 250,831
----------------------------------------------------------------------
Source: IRS' AIMS file on tax year 1992 closed audit cases.
As shown in table I.1, the 218,467 taxpayers with tax assessments of
more than $500 potentially meet the eligibility criteria for innocent
spouse relief. Thus, this is the maximum number of taxpayers that
could request innocent spouse relief because of an audit assessment.
ESTIMATED NUMBER OF INNOCENT
SPOUSE CASES THAT COULD RESULT
FROM THE UNDERREPORTER PROGRAM
--------------------------------------------------------- Appendix I:2
Of the 1.8 million tax year 1992 underreporter cases, 89 percent
(about 1.6 million) were assessed additional taxes in fiscal year
1995. The remaining 11 percent (about 200,000 cases) were assessed
additional taxes in either 1994 or 1996.
Since almost 90 percent of 1992 underreporter cases were assessed in
1995, we estimated the number of potential tax year 1992 innocent
spouse cases by evaluating a stratified probability sample of 463
underreporter cases assessed in 1995. Based on the tax year 1992
returns in this sample, we estimated how many of the 1.6 million tax
year 1992 returns assessed in 1995 were "married, filing joint"
returns and, of those, how many had assessments under $500 and
assessments of $500 or more. Although we did not have data to make
similar estimates for the 200,000 tax year 1992 returns assessed in
1994 and 1996, we assumed that the filing status and assessed amounts
for these returns would be similar to the returns in the 1995 sample
where the assessments occurred either 2 years or 4 years after the
tax year. We could not directly test this assumption, but our
analysis of the data indicates that various characteristics of the
returns, such as the amount assessed and the complexity of the
underreporting problem, determined when IRS processed the case. We
weighted the sample of 463 underreporter cases to represent the
number of 1992 tax year returns processed in 1995 and other years and
then estimated the potential innocent spouse cases for tax year 1992.
We accounted for the stratified sample design and unequal weights for
the sample cases when calculating the sampling errors.
The point estimates and sampling errors for 1992 are given in table
I.2.
Table I.2
Estimates of Tax Year 1992 Taxpayers Who
Could Be Eligible for Innocent Spouse
Relief Based on Underreporter
Assessments
Point estimates
and sampling
Data used to calculate underreporter cases errors
-------------------------------------------------- ------------------
Estimated number of tax year 1992 assessments for 993,536 � 150,469
"married, filing joint" returns
Estimated number of tax year 1992 joint filer 624,734 � 151,703
cases with assessments of $500 or less
Estimated number of tax year 1992 joint filer 368,802 � 55,822
cases with assessments of more than $500
----------------------------------------------------------------------
Source: GAO's analysis of IRS' underreporter data
Adding the tax year 1992 joint filer underreporter cases with tax
assessments of more than $500 in table I.2 to the 218,467 joint filer
audit cases in table I.1 results in an estimated potential universe
of innocent spouse cases of approximately 587,000 that is bounded by
a confidence interval of between 531,447 and 643,091 returns.
Additional uncertainty about this estimate comes from our previously
mentioned assumption about the characteristics of the 11 percent of
the 1992 returns that were not assessed in fiscal year 1995.
ESTIMATED NUMBER OF MARRIED
TAXPAYERS WHO DIVORCE EACH YEAR
--------------------------------------------------------- Appendix I:3
To estimate the number of taxpayers who divorce each year, we
obtained data from the Bureau of the Census and the Department of
Health and Human Services. Census' report, entitled Marital Status
and Living Arrangements: March 1993, shows that there were
114,602,000 married individuals in the United States at that time. A
Monthly Vital Statistics Report, dated October 1995 and prepared by
the National Center for Health Statistics within the Department of
Health and Human Services, states that approximately 1,191,000
divorces were granted in the United States in 1994. Therefore, an
estimated 2,382,000 individuals who were married in 1993 had divorced
in 1994. Dividing this number of individuals by the total number of
married individuals in 1993 results in an annual divorce rate of
2.0784 percent, which we have rounded to 2 percent for the
calculations in our study. To estimate how many of the potential
innocent spouse cases would involve divorced couples over a 3-year
period, we multiplied 583,410 by the annual divorce rate of 2 percent
and then multiplied by 3.
INFORMATION ON PROBLEM RESOLUTION
OFFICE INNOCENT SPOUSE CASES
========================================================== Appendix II
CASES BETWEEN JANUARY AND JUNE
1996
-------------------------------------------------------- Appendix II:1
As part of IRS' efforts to address the problems of divorced and
separated taxpayers, the headquarters Problem Resolution Office asked
the Problem Resolution Offices at four district offices and one
service center to identify all cases where the taxpayer raised the
issue of innocent spouse relief between January and June 1996. The
offices identified 51 cases and forwarded them to headquarters, where
we reviewed them.
Of the 51 cases, 20 were not innocent spouse cases--for example, the
taxpayer was actually an injured spouse\13 or was requesting relief
for taxes reported as owed on the original return. Of the remaining
31 cases, 4 involved taxpayers who had already been granted innocent
spouse relief but IRS had not correctly transferred the tax liability
to the culpable spouse. We did not include these cases since the
files contained information only on the processing problem, not the
innocent spouse issues. Table II.1 describes the characteristics of
the remaining 27 cases.
Table II.1
Characteristics of Problem Resolution
Cases
Number of
cases with
Characteristic Average Range data\a
------------------------------------ -------- ---------- ----------
Years elapsed\b 7 years 2 years 27
to
17 years
Amount of assessment $12,000 $900 to 23
$50,000
Annual income $37,000 $7,000 to 13
$77,000
----------------------------------------------------------------------
\a The Problem Resolution Office case files did not always include
all these data .
\b Years elapsed between the tax year assessed and the taxpayer's
contact with the Problem Resolution Office.
Source: IRS' Problem Resolution Office case files.
For the 24 cases where we could find data, all of the taxpayers began
requesting help because of collection activity, such as having their
wages levied by IRS. For the 24 cases where we could determine
marital status, 20 taxpayers were divorced and 4 were separated. For
the 21 cases where a decision had been reached, IRS granted relief in
10 of the cases.
--------------------
\13 Injured spouses are joint filers whose joint refunds have been
seized to pay certain of their spouses' nontax debts, such as
past-due child or spousal support or a federal loan.
CASE HISTORIES
-------------------------------------------------------- Appendix II:2
The following cases histories illustrate some of the issues involved
in innocent spouse cases.
CONFUSION OVER PROCEDURES
AND DOCUMENTING DEDUCTIONS
------------------------------------------------------ Appendix II:2.1
According to IRS' records, a taxpayer learned of an assessment of
over $3,000 against a 1985 joint return when IRS levied her wages in
1992. The assessment was generated primarily by her ex-husband's
disallowed business and moving expenses, although he also had some
unreported income. The taxpayer contacted the Problem Resolution
Office about innocent spouse relief. Problem Resolution Office staff
initially could not explain how to apply for relief but provided
assistance during subsequent contacts. The taxpayer submitted
documentation demonstrating that the unreported income was generated
by her husband and received relief for about $200. According to an
IRS official, she could not substantiate her husband's disallowed
business expenses and was held liable for the remainder of the tax.
IMPACT ON NEW SPOUSES AND
ORIGINAL RETURN
------------------------------------------------------ Appendix II:2.2
A taxpayer whose ex-husband was a wanted fugitive had outstanding tax
liabilities because her ex-husband had failed to report income on
their joint return for 1 tax year and had not paid the tax reported
for 2 subsequent tax years. The taxpayer remarried, and IRS placed
liens against her new husband's property. IRS denied innocent spouse
relief, in part because the liability for 2 years was for taxes
reported as due on the original return. IRS did accept an Offer in
Compromise for all 3 years.
RELIEF ON ORIGINAL RETURN
AND KNOWLEDGE TEST
------------------------------------------------------ Appendix II:2.3
In 1995, a taxpayer wrote to IRS to protest a balance due on joint
returns for 3 tax years. The taxes were attributable to income
derived from her ex-husband's fraudulent activities. In 1996, IRS
informed the taxpayer she was not eligible for innocent spouse relief
for 2 tax years because these balances were for taxes reported as due
on the original returns but not paid when the returns were filed.
However, IRS staff informed the taxpayer they would consider innocent
spouse relief for 1 year if the taxpayer could demonstrate she had no
knowledge of the unreported income.
IRS denied her request for innocent spouse relief because she could
not meet the knowledge test. Subsequently, she submitted third-party
statements that she did not live a lavish or enhanced lifestyle as
well as copies of police records on her ex-husband's arrest and
trial. IRS eventually granted innocent spouse relief for that 1
year.
KNOWLEDGE TEST
------------------------------------------------------ Appendix II:2.4
A taxpayer learned of an assessment of over $12,000 on a joint 1991
return when IRS levied her wages. The couple had legally separated
and sold their residence. Although she had reinvested her half of
the capital gain on the sale, her ex-husband had not done so within
the 24 months required to defer taxes. IRS denied innocent spouse
relief because the taxpayer could not meet the knowledge test.
According to IRS, she knew of the possibility of the additional tax
when she signed the return.
KNOWLEDGE TEST, DIVORCE
DECREE, DOLLAR THRESHOLD
------------------------------------------------------ Appendix II:2.5
A taxpayer learned of an assessment of about $1,200 on joint returns
for 2 years when IRS seized her 1995 tax refund. The assessment was
generated by her ex-husband's unreported income. The taxpayer argued
that the couple had maintained separate checking and savings
accounts, and therefore she did not know of the unreported income.
Furthermore, the divorce decree specified that she would assume
outstanding credit card debts and her ex-husband would be responsible
for all other debts incurred during the marriage. IRS denied
innocent spouse relief for 1 year because the assessment for that
year was less than the $500 threshold. IRS denied innocent spouse
relief for the other year because the taxpayer did not meet the
knowledge requirement. Because the unreported income was more than
75 percent of the ex-husband's total income, IRS staff believe she
should have been aware of the income earned even though the spouses
had separate accounts.
DOLLAR THRESHOLDS
------------------------------------------------------ Appendix II:2.6
A taxpayer was assessed over $3,000 on joint returns filed in 4 tax
years generated by her husband's disallowed deductions for gambling
losses. She was denied innocent spouse relief for 1 year because the
assessment for that year was less than the $500 threshold. She was
denied innocent spouse relief for the other 3 years because the
assessment in each of those years was less than 25 percent of her
adjusted gross income.
STATE EFFORTS TO ADDRESS JOINT AND
SEVERAL LIABILITY ISSUES
========================================================= Appendix III
We contacted the state tax agencies to discuss how they handle
various issues surrounding joint and several liability when
administering the state income tax system. Seven states--Alaska,
Florida, Nevada, South Dakota, Texas, Washington, and Wyoming--do not
tax personal income. Since the tax agencies for these states do not
administer any joint income tax returns, we excluded them from our
review. The remaining 43 states and the District of Columbia do
administer joint returns; however 2 states--New Hampshire and
Tennessee--tax only interest and dividends.
Table III.1 provides information by state on the type of liability
standard the states apply to joint returns, the type of innocent
spouse relief they administer, and whether they are bound by divorce
decrees.
Table III.1
State Liability Standards, Innocent
Spouse Provisions, and Use of Divorce
Decrees\a
Joint return Innocent spouse
with relief based on Less restrictive
proportionate Internal Revenue innocent spouse Bound by divorce
State liability Code relief decrees
-------- ---------------- ---------------- ---------------- ----------------
Alabama X
Arizona X
Arkansas
Californ X X
ia
Colorado
Connecti
cut
Delaware X
District X
of
Columbia
Georgia X
Hawaii X
Idaho X
Illinois X
Indiana
Iowa X
Kansas
Kentucky X
Louisian X
a
Maine X
Maryland X
Massachu X
setts
Michigan X
Minnesot X
a
Mississi
ppi
Missouri
Montana
Nebraska X
New
Hampshir
e
New
Jersey
New
Mexico
New York X
North X
Carolina
North X
Dakota
Ohio
Oklahoma X
Oregon X
Pennsylv
ania
Rhode
Island
South X
Carolina
Tennesse
e
Utah X
Vermont
Virginia
West
Virginia
Wisconsi X X
n
================================================================================
Total 1 18 6 3
--------------------------------------------------------------------------------
\a Alaska, Florida, Nevada, South Dakota, Texas, Washington, and
Wyoming do not tax personal income and we excluded them from our
review.
ESTIMATED COSTS FOR PROCESSING
SEPARATE OR MODIFIED JOINT RETURNS
========================================================== Appendix IV
This appendix describes how we estimated the increased
returns-processing costs IRS would face if a proportionate liability
standard were administered by either (1) eliminating joint returns
and requiring all taxpayers to file separately; or (2) retaining
joint returns but modifying them so that each spouse's income,
deductions, and credits are reported in separate columns.
ESTIMATED COST OF PROCESSING
SEPARATE RETURNS
-------------------------------------------------------- Appendix IV:1
To estimate the cost of processing up to 48 million additional
returns, we determined the distribution of the returns among the
standard forms 1040, 1040A, and 1040PC in the 1992 Statistics of
Income data. We then estimated the cost of processing these returns
using IRS' Document 6746, Cost Estimate Reference for Service Center
Returns Processing for Fiscal Year 1994. The calculation is shown in
table IV.1.
Table IV.1
Estimated Cost of Processing an
Additional 48 Million Tax Returns
Returns
processing
Number cost per
Type of return (thousands) thousand Cost
------------------------- ------------- ------------- -------------
1040 37,802 $4,231.89 $159,973,906
1040A 8,092 3,690.48 29,863,364
1040PC 2,126 4,231.89 8,996,998
======================================================================
Total $198,834,268
----------------------------------------------------------------------
Source: IRS' 1992 SOI data and Document 6746, Cost Estimate
References for Service Center Returns Processing for Fiscal Year
1994.
ESTIMATED COST OF PROCESSING
MODIFIED RETURNS
-------------------------------------------------------- Appendix IV:2
To estimate the additional data entry costs for 48 million tax
returns if the returns were modified to show two income streams, we
also used the distribution of the returns among the standard forms
1040, 1040A, and 1049PC in the 1992 Statistics of Income data and
IRS' Cost Estimate Reference for Service Center Returns Processing
Fiscal Year 1994. The calculation is shown in table IV.2.
Table IV.2
Estimated Additional Data Entry Costs
for Modified Joint Returns
Data entry
Number cost
Type of return (thousands) per thousand Cost
------------------------- ------------- ------------- -------------
1040 37,802 $412.64 $15,598,617
1040A 8,092 263.54 2,132,566
1040PC 2,126 412.64 877,273
======================================================================
Total $18,608,456
----------------------------------------------------------------------
Source: IRS' 1992 SOI data and Document 6746, Cost Estimate
References for Service Center Returns Processing for Fiscal Year
1994.
DISTRIBUTION OF JOINT, SEPARATE,
AND UNKNOWN INCOME FOR 1992 JOINT
INCOME TAX RETURNS
=========================================================== Appendix V
The extent to which adopting a proportionate liability standard would
create an administrative burden for IRS depends partly on how easily
IRS could use an automated reporting system to allocate income and
mortgage interest payments between the two spouses. We reviewed a
stratified probability sample of 200 joint tax returns selected from
IRS' 1992 Statistics of Income database to determine how much income
was reported separately for each spouse or jointly for the married
couple. We projected the results to the universe of 48 million
couples who filed using the "married, filing jointly" status at a 95
percent confidence level.
Of all the income in our sample, 77 percent was reported separately.
About 2 percent of the income was reported as joint income, and we
could not determine whether the income was separate or joint for 12
percent. As shown in table V.1, some types of income were generally
reported separately. For example, IRA distributions and unemployment
compensation were always reported separately, as was most farm income
(Schedule F). As a result, IRS would have little difficulty
allocating this income under a proportionate liability standard.
In contrast, state and local tax refunds are generally reported in
both spouses' names. As a result, IRS could have difficulty
allocating this income between the two taxpayers.
We had difficulty determining whether other income categories were
separate or joint. For example, we generally could not determine
which taxpayer earned the profit or loss reported on Schedule E,
supplemental net gains or losses, or net operating losses.
Table V.1
Percentage of Income Reported Jointly,
Separately, or Could Not Determine
Separate Could not
Income source Joint income income determine Total\a
------------------------ ------------ ------------ ------------ ============
Wages, salaries, tips 0% 96�2 % 4�2% 100%
Taxable interest income 24�10 36�12 40�17 100
Tax-exempt interest 0 0 100 100
income
Dividend income 26�17 68�17 5�3 100
Taxable state and local 82�15 4�4 15�13 100
tax refunds
Schedule C Net Income or 0 82�29 18�29 100
Loss
Supplemental net gains 0 0 100�.1 100
or losses
Taxable IRA 0 100 0 100
distributions
Taxable pensions and 0 100�.1 0 100
annuities
Schedule E Profit or 9�14 4�5 87�16 100
Loss
Schedule F Profit or 0 97�5 3�5 100
Loss
Unemployment 0 100 0 100
compensation
Taxable Social Security 0 83�21 17�21 100
benefits
Net operating loss 0 0 100 100
--------------------------------------------------------------------------------
\a Rows may not total due to rounding.
Source: IRS, 1992 Statistics of Income database.
(See figure in printed edition.)Appendix VI
COMMENTS FROM THE INTERNAL REVENUE
SERVICE
=========================================================== Appendix V
(See figure in printed edition.)
MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix VII
GENERAL GOVERNMENT DIVISION,
WASHINGTON, D.C.
Ralph Block, Assistant Director, Tax Policy and Administration Issues
Nancy Peters, Assignment Manager
Elizabeth Scullin, Communications Analyst
SAN FRANCISCO/SEATTLE FIELD OFFICE
Jack Erlan, Evaluator-in-Charge
Jonda Van Pelt, Senior Evaluator
*** End of document. ***