Internal Revenue Service: Challenges Remain in Combating Abusive Tax Shelters

Published by the Government Accountability Office on 2003-10-21.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                             United States General Accounting Office

GAO                          Testimony
                             Before the Committee on Finance, U.S. Senate

For Release on Delivery
Expected at 10:00 a.m. EDT
Tuesday, October 21, 2003
                             INTERNAL REVENUE
                             Challenges Remain in
                             Combating Abusive Tax
                             Statement of Michael Brostek
                             Director, Tax Issues



                                                October 21, 2003

                                                INTERNAL REVENUE SERVICE

                                                Challenges Remain in Combating
Highlights of GAO-04-104T, a testimony to       Abusive Tax Shelters
the Committee on Finance, U.S. Senate

Recent scandals involving                       By their nature, abusive tax shelters are varied, complex, and difficult to
corporations, company executives,               detect and measure. Abusive shelters manipulate many parts of the tax code
and accounting, law, and                        or regulations and may involve steps to hide the transaction within a tax
investment banking firms                        return. In recent years, IRS has been accumulating information about them
heightened awareness of abusive                 and, although it does not have a reliable measure of the size of the abusive
tax shelters and highlighted the
importance of the Department of
                                                shelter problem, has come to believe that abusive shelters deserve
the Treasury and the Internal                   substantially increased attention. IRS continues to gather more information
Revenue Service (IRS) addressing                to better define the scope of the problem and has data sources, all with their
them. During 1999, Treasury                     own limitations, that suggest abusive tax shelters total tens of billions of
issued a report indicating that                 dollars of potential tax losses over about a decade.
abusive shelters were a large and
growing problem, involving billions             IRS’s broad-based strategy for addressing abusive shelters included:
of dollars of tax reductions.
Treasury was concerned that                     •   targeting promoters to head off the proliferation of shelters;
abusive shelters could ultimately               •   making efforts to deter, detect, and resolve abuse;
undermine the integrity of the
voluntary compliance tax system.
                                                •	  offering inducements to individuals and businesses to disclose their use
                                                    of questionable tax practices; and
GAO’s statement today is based on               • 	 using performance indicators to measure outputs and some outcomes
work done at the request of the                     and intending to go down the path it has started and develop long-term
Chairman and the Ranking Minority                   performance goals and measures linked to those goals. Without these
Member of the Senate Committee                      latter elements, Congress would find gauging IRS’s progress difficult.
on Finance to examine IRS’s
strategy for dealing with abusive               In allocating resources to shelters, IRS used a systematic decision-making
tax shelters. In reporting on                   process that relied on admittedly limited information. It planned to shift
abusive shelters, GAO is describing             significant resources in fiscal years 2003 and 2004 to address abusive
                                                shelters but faces challenges, especially in the near term, in addressing
•    their nature and scope,
•	   IRS’s strategy and enforcement
                                                abusive shelters due to a growing workload and limited information on how
     mechanisms to combat them                  long it takes to examine shelter cases. IRS’s understanding of how many
     and the performance goals and              staff will be needed to address the problem over what period will continue to
     measures IRS uses to track its             evolve as it gains a better understanding of the problem’s scope.
     major effort in that area, and             Shift in Examination Resources Allocated to Abusive Tax Shelters
•	   the decision-making process
     IRS used and the plans it has
     to devote more resources to
     addressing abusive shelters.


To view the full product, including the scope   Note: Fiscal year 2002 full-time equivalents include actual time spent on the entire returns
and methodology, click on the link above.       containing shelters, not on the shelter issues alone. Fiscal year 2003 and 2004 full-time equivalents
For more information, contact Michael           are planned amounts that are focused more on the shelter issues themselves.
Brostek at (202) 512-9110 or
Mr. Chairman and Members of the Committee:

I appreciate the opportunity to testify on the Internal Revenue Service’s
(IRS) efforts to deal with abusive tax shelters. I am using the term “abusive
shelters” to describe very complicated transactions promoted to
corporations and wealthy individuals to exploit tax loopholes and provide
large, unintended tax benefits. Recent scandals involving corporations,
company executives, and accounting, law, and investment banking firms
heightened awareness of abusive shelters and highlighted the importance
of the Department of the Treasury and IRS addressing the problem. During
1999, Treasury issued a report indicating that abusive shelters were a large
and growing problem, involving billions of dollars of tax reductions.1
Treasury was concerned that abusive shelters could ultimately undermine
voluntary compliance by eroding the integrity of the tax system. In
response to information pointing to the rapid growth of abusive shelters,
IRS formalized a strategic initiative in fiscal year 2000 to strengthen its
capacity to deal with abusive corporate shelters. One element of IRS’s
initiative involved creating a central office within the Large and Mid-Size
Business (LMSB) Division to coordinate and guide efforts to curb the
growth of abusive shelters.

My statement today is based on work we have done at the request of the
Chairman and the Ranking Minority Member. In examining abusive
shelters, we focused on (1) their nature and scope, (2) IRS’s strategy and
enforcement mechanisms to combat them and the performance goals and
measures IRS uses to track its major effort in that area, and (3) the
decision-making process IRS used to allocate resources to abusive shelters
and the plans it has to devote more resources to addressing abusive
shelters. We were also asked to provide information on IRS’s Schedule K-1
document matching program, which we are including in appendix I.

To do our work, we

 Department of the Treasury, The Problem of Corporate Tax Shelters (Washington, D.C.: July

Page 1                                                                       GAO-04-104T
•	 analyzed IRS’s and other shelter reports, publications, data, and other
   documentation providing insight into the characteristics, complexity,
   size, and type of the problem;2

•	 reviewed IRS’s planning documents with information on its strategies,
   measures, and resources;

•	 compared the contents of IRS’s planning documents to Government
   Performance and Results Act of 1993 (GPRA)3 criteria for what elements
   strategic planning should include; and

•	 interviewed agency officials about their views on, among other things,
   the problem’s nature and scope and IRS’s strategy.

We did our work from September 2002 through August 2003 in accordance
with generally accepted government auditing standards. As agreed, we are
also discussing the related problem of abusive tax schemes in a report to be
released in the near future. Abusive tax schemes are used more by
individuals than by large businesses and encompass such distortions of the
tax system as falsely describing the law (saying, for example, that the
income tax is unconstitutional), misrepresenting facts (for instance,
promoting the deduction of personal expenses as business expenses), and
using trusts or offshore bank accounts to hide income. The boundary
between what we are calling an abusive tax shelter and an abusive scheme
is not always clear. Organizationally, although IRS’s LMSB Division has
lead responsibility for combating abusive shelters, abusive shelters are
pursued by IRS’s Small Business/Self-Employed Division when they are
used by businesses with assets of less than $10 million or by high-wealth
individuals with complicated tax returns.

My statement today will make the following points:

 As part of this work, we tested the tax shelter database maintained by the Office of Tax
Shelter Analysis (OTSA) by reviewing related documentation, interviewing knowledgeable
agency officials, and doing electronic testing, finding that the required data elements were
sufficiently reliable for the purposes of our work. This finding does not mean, however, that
the database contains all the information that would be needed to estimate the full size of
the abusive shelter problem.
 Pub. L. No.103-62.

Page 2                                                                         GAO-04-104T
•	 By their nature, abusive shelters are varied, complex, and difficult to
   detect and measure. Abusive shelters manipulate many parts of the tax
   code or regulations and may involve steps to hide the transaction within
   a tax return. In recent years, IRS has been accumulating information
   about abusive shelters and the extent that they were promoted, and it
   has come to believe that abusive shelters deserve substantially
   increased attention. Suffice it to say, although they do not have a
   reliable measure of the size of the abusive shelter problem, Treasury and
   IRS believe that tens of billions of dollars of taxes are being improperly
   avoided and the potential for the proliferation of abusive shelters is
   strong. IRS continues to gather more information to better define the
   scope of the problem and has several data sources, each with certain
   limitations, that point to billions in tax losses. As of September 30, 2003,
   a database on shelter transactions that IRS has publicly declared to be
   tax avoidance transactions suggested the potential tax loss to be about
   $33 billion, the majority of which was concentrated from tax year 1993
   through the present. This database included only transactions disclosed
   to or discovered by IRS. In addition, an IRS contractor estimating
   annual tax gaps resulting from abusive shelters estimated that the
   annual average of foregone taxes between 1993 and 1999 could have
   been as small as about $11.6 billion or as large as about $15.1 billion.
   However, Treasury, IRS, the contractor, and we all have concerns about
   the reliability of the contractor’s estimates because of methodological
   and data constraints that the contractor faced.

•	 The broad-based strategy reflected in IRS planning documents included
   various features as well as elements of strategic planning:

   • targeting promoters to head off the proliferation of shelters;

   • making efforts to deter, detect, and resolve abuse;

   • coordinating efforts throughout IRS;

   •	 offering inducements to individuals and businesses to disclose their
      use of questionable tax practices; and

   •	 using performance indicators to measure outputs and some
      outcomes and intending to continue down the path it has started and
      develop long-term performance goals and measures linked to those
      goals. Without these latter elements, Congress would find gauging
      IRS’s progress difficult.

Page 3                                                             GAO-04-104T
              •	 In developing this strategy, IRS has had to make decisions about staffing
                 allocations and what can be accomplished on the basis of admittedly
                 limited information. After using a systematic process to determine
                 staffing priorities, IRS planned a significant shift in resources to address
                 abusive shelters in fiscal years 2003 and 2004. However, it faces
                 challenges, especially in the near term, in addressing abusive shelters
                 due to a growing workload and limited information on how long it takes
                 to examine shelter cases. IRS’s understanding of how many staff will be
                 needed to address the problem over what period will continue to evolve
                 as IRS gains a better understanding of the problem’s scope.

Background	   Although IRS has no single, authoritative definition of abusive shelters, IRS
              generally characterizes abusive shelters as very complicated transactions
              that sophisticated tax professionals promote to corporations and wealthy
              individuals, exploiting tax loopholes and reaping large and unintended tax
              benefits. As the Joint Committee on Taxation has said, “taxpayers and tax
              administrators have struggled in determining the line between legitimate
              ‘tax planning’ and unacceptable ‘tax shelters.’” Even though, it continued,
              “there is no uniform standard as to what constitutes a tax shelter … there
              are statutory provisions, judicial doctrines, and administrative guidance
              that attempt to limit or identify transactions in which a significant purpose
              is the avoidance or evasion of income tax.”4

              Abusive shelters have been promoted by some accounting firms, law firms,
              and investment banks. Investors in these abusive shelters range from large
              and small corporations to wealthy individuals. IRS approaches the tax
              shelter enforcement problem from both the promoter and investor
              perspectives. IRS promoter investigations are designed to learn (1) what
              abusive shelters have been promoted, if the shelters are registered,5 and
              possibly how much they cost investors, (2) who purchased the shelters and
              what tax savings the investors expect, and (3) whether promoters should
              pay penalties for their activities. IRS examines investor and other tax

               Joint Committee on Taxation, Background and Present Law Relating to Tax Shelters, JCX-
              19-02 (Washington, D.C.: Mar. 19, 2002).
               A promoter or other tax shelter organizer must register a tax shelter with the Secretary of
              the Treasury by describing it and its tax benefits. The Secretary assigns the shelter an
              identification number.

              Page 4                                                                          GAO-04-104T
returns to see if income, expenses, taxes, and credits are accurately

In a June 2002 letter, Treasury responded to congressional questions about
whether Treasury had a comprehensive strategy for combating tax
avoidance. In his letter to the then Ranking Member of the Committee on
Finance, then Secretary of the Treasury O’Neill addressed the actions being
taken to combat abusive shelters, referring to Treasury’s March 20, 2002,
enforcement proposals on the topic. The proposals said that IRS had made
significant organizational improvements to coordinate its response to
ongoing abusive tax shelters. Treasury, all of IRS’s operating divisions, and
IRS’s Office of Chief Counsel are involved in combating abusive shelter

Within IRS, LMSB has primary responsibility for combating abusive tax
shelter activity. LMSB’s OTSA was created in February 2000 to centralize
and coordinate the IRS response nationwide. As shown in figure 1, OTSA is
the focal point for IRS shelter activities, overseeing promoter tax shelter
registrations; taxpayer disclosures of tax shelters; hotline tip analysis and
referral; and issue coordination and interface between the Office of Chief
Counsel, Treasury, the Tax Shelter Committee, the 6700 Committee
(referring to section 6700 of the Internal Revenue Code), and external
stakeholders.6 The Tax Shelter Committee oversees LMSB’s tax shelter
program. The committee is composed of the Commissioner and Deputy
Commissioner of LMSB, the Director of Pre-Filing and Technical Guidance,
LMSB Division Counsel, five Industry Directors, the Director of
International, and the Directors of Field Specialists and Research and
Program Planning. The 6700 Committee serves under the Tax Shelter
Committee and approves all LMSB tax shelter promoter activities. The
financial services’ industry director chairs this committee. IRS’s appeals
function receives and evaluates taxpayer objections to IRS examination
determinations and may agree with those determinations or reduce or
eliminate changes to tax returns resulting from them. The Office of Chief
Counsel plays an integral role in combating shelters through summons
enforcement and targeted litigation. By litigating, IRS establishes case law
supporting IRS enforcement programs and aims to diminish the incentives
taxpayers find for investing in tax avoidance transactions by increasing the
risks and costs of IRS discovery.

 Section 6700 covers penalties for promoters of abusive shelters.

Page 5                                                              GAO-04-104T
                         Figure 1: OTSA’s Role in Coordinating IRS Work on Abusive Shelters

                                                                                Tax Shelter
                                                                        oversees LMSB’s tax shelter
                                                                          program, and its affiliated
                                                                         6700 Committee approves
                                                                               LMSB promoter
                                  External stakeholders                                                       Counsel
                                  such as Treasury, tax                                                  works on guidance,
                               practitioners, and the press,                                            helps develop issues,
                                  provide guidance and                                                       and litigates
                                    receive information
                                                                             is the focal point for
                                                                            IRS shelter activities

                                      Other internal
                                                                                                          IRS field offices
                                    for instance, IRS
                                                                                                          conduct shelter
                               operating divisions, such as
                                SB/SE, provide feedback
                                      and guidance
                                                                               handles shelter

                         Source: Compiled by GAO from IRS information

Nature of Abusive        Abusive shelters are complex transactions that manipulate many parts of
                         the tax code or regulations and are typically buried among “legitimate”
Shelters Is Varied and   transactions reported on tax returns. Because these transactions are often
Complex                  composed of many pieces located in several parts of a complex tax return,
                         they are essentially hidden from plain sight, which contributes to the
                         difficulty of determining the scope of the abusive shelter problem. Often
                         lacking economic substance or a business purpose other than generating
                         tax benefits, abusive shelters are promoted by some tax professionals,
                         often in confidence, for significant fees, sometimes with the participation
                         of tax-indifferent parties, such as foreign or tax-exempt entities. They may
                         involve unnecessary steps and flow-through entities, such as partnerships,
                         which make detection of these transactions more difficult.

                         When a transaction has certain abusive characteristics defined by section
                         6111 of the Internal Revenue Code, the promoter or other tax shelter

                         Page 6                                                                                         GAO-04-104T
organizer is required to register it, describing the transaction and its tax
benefits to the Secretary of the Treasury. This registration requirement
enables Treasury and IRS to identify and evaluate questionable
transactions. Under recently issued Treasury regulations,7 effective
February 28, 2003, there are six categories of transactions for which
promoters must maintain lists of investors who have entered into the
transactions, and investors must disclose the transactions into which they
have entered. The rules are designed to allow IRS to use information from
investors to identify promoters who do not register transactions and to use
promoter registrations and investor lists to identify investors who fail to
disclose transactions. The six categories are

• transactions offered under conditions of confidentiality,

• transactions including contractual protections to the investor,

• transactions resulting in specific amounts of tax losses,

•	 transactions generating a tax benefit when the underlying asset is held
   only briefly,

•	 transactions generating differences between financial accounts and tax
   accounts greater than $10 million, and

• “listed transactions.”

A “listed transaction” is a transaction that is the same as or similar to one of
the types of transactions IRS has determined to be a tax avoidance
transaction. For a transaction to be a listed transaction, IRS must issue a
notice, regulation, or other form of published guidance informing taxpayers
of the details of the transaction. As of mid-August 2003, IRS had listed 27
kinds of abusive tax shelter transactions, a number that, as figure 2 shows,
has grown more quickly in recent years than it had grown earlier.

 Treas. Reg. Sec. 301.6112-1 and Treas. Reg. Sec. 1.6011-4.

Page 7                                                              GAO-04-104T
Figure 2: Cumulative Number of Listed Transactions over Time, 1990–mid-August 2003, and Transaction Descriptions

          1990                                      1999                                     2001                                    2003
          Revenue Ruling 90-105: Certain            Treasury Regulation Section              Notice 2001-16: Intermediary            Revenue Ruling 2003-6:
          accelerated deductions                    1.643(a)-8: Certain distributions        transactions                            Abuses associated with S
                                                    from charitable remainder trusts         Notice 2001-17: Section 351             corporation employee stock
          1995                                                                               contingent liability                    option plans
                                                    Revenue Ruling 99-14: Lease-
          Notice 95-34: Certain employer            in/lease-out or LILO transactions                                                Notice 2003-22: Offshore
                                                                                             Notice 2001-45: Section 302
          welfare fund trusts                                                                                                        deferred compensation
                                                    Notice 99-59: Bond and option            basis-shifting transactions
          Notice 95-53: Lease strips;                                                                                                arrangements
                                                    sales strategy (BOSS) transactions
          superseded by Notice 2003-55                                                       2002                                    Notice 2003-24: Certain
                                                    Treasury Regulation Section              Notice 2002-21: Inflated basis          welfare benefit fund trusts
          1998                                      1.7701(l)-3: Fast pay or step-down       custom adjusted rate debt
                                                    preferred transactions                                                           Notice 2003-47:
          Notice 98-5, part II: Foreign tax                                                  transactions                            Compensatory stock options
          credit transactions
                                                    2000                                     Notice 2002-35 and Revenue              Notice 2003-54: Common
          Contingent installment sale               Revenue Ruling 2000-12: Debt             Ruling 2002-30: Notional                trust fund straddle
          transactions                              straddles                                principal contracts

                                                    Notice 2000-44: Inflated                 Revenue Ruling 2002-46 and
                                                    partnership basis transactions           Revenue Ruling 2002-73:
    Number                                          (son of BOSS)                            Section 401k accelerators
    30                                              Notice 2000-60: Stock                    Notice 2002-50: Partnership
                                                    compensation transactions                straddle tax shelter
                                                    Notice 2000-61: Guam trust               Notice 2002-65: Pass-through
    25                                                                                       entity straddle tax shelter
                                                                                             Notice 2002-70: Certain
                                                                                             reinsurance arrangements




            1990        1991        1992         1993      1994         1995         1996   1997      1998       1999         2000   2001         2002        2003

 Source: Compiled by GAO from IRS information.

                                                             Disputes between IRS and taxpayers about the abusive nature of a
                                                             transaction may be litigated. In some, but not all, cases, the courts have
                                                             upheld the government position. The following cases illustrate features of
                                                             abusive shelters:

                                                             Page 8                                                                                        GAO-04-104T
•	 In 1993, a corporation began a company-owned life insurance (COLI)
   program in which the company purchased whole-life insurance on
   36,000 employees for which the company was the sole beneficiary. The
   company then borrowed money against the policies at interest rates that
   averaged 11 percent and deducted the interest expense and
   administrative fees from income on its tax returns. Over 60 years, the
   interest costs and administrative fees would have exceeded the cash
   surrender value of the policies and benefits paid by several billion
   dollars. IRS disallowed the deductions and the case was litigated.
   Despite the fact that the money the company made on this arrangement
   may have been used to fund the company’s benefits program, or for
   other business purposes, the court found that the function of the
   program itself was only to generate tax deductions. As a result, the Tax
   Court sustained the IRS disallowance of deductions and concluded that
   the COLI program was a sham.8 The Eleventh Circuit Court of Appeals
   affirmed the Tax Court’s decision.

•	 A company had a sizable gain from the sale of a subsidiary and wanted
   to avoid or minimize paying tax on the gain. An investment bank
   proposed forming an offshore partnership with a foreign corporation (a
   tax-indifferent party) for the express purpose of sheltering the capital
   gains of its corporate client. The partnership purchased and quickly
   resold notes in a contingent installment sale transaction. The
   partnership earned a large capital gain, most of which it allocated to the
   foreign corporate partner. Later, related losses were allocated to the
   U.S. corporation, generating an approximate $100 million capital loss for
   the investment bank’s client. The corporation used this capital loss to
   shelter its U.S.-based capital gains. Both the Tax Court and the Third
   Circuit Court of Appeals ruled that the transaction lacked economic
   substance.9 The Third Circuit, in addition to requiring economic
   substance, held that a transaction must have a subjective nontax
   business motive to be respected for tax purposes.10 For this transaction,
   the investment bank was to earn a fee of $2 million. This was one of 11

Winn-Dixie Stores, Inc. v. Commissioner, 113 T.C. 254 (1999), aff’d, 254 F. 3d 1313 (11th
Cir. 2001).
 ACM Partnership v. Commissioner, 157 F. 3d 231 (3d Cir. 1998), aff’g, 73 T.C.M. 2189
(1997), cert. denied, 526 U.S. 1017 (1999).
     Id. at 248.

Page 9                                                                        GAO-04-104T
                               such partnerships formed over a 1-year period from 1989 to 1990 by the
                               investment bank.

Several Sources           IRS has information that suggests the scope of abusive shelters totaled tens
                          of billions of dollars over about a decade,11 but those estimates are based
Indicate That the         on limited data. This information comes from an OTSA database,
Scope of Abusive          examinations of large corporations, and a contractor study. Information
                          contained in the OTSA database includes transactions disclosed to or
Shelters Is in the Tens   discovered by IRS and estimates of potential tax losses. The tax loss
of Billions of Dollars,   estimates vary from being IRS officials’ recommended taxes based on
Though All Are Based      examining some transactions to taxpayer judgments regarding potential
                          losses in cases where examinations have not been done. In addition to
on Limited Data           being based on judgments, the database does not include any reductions
                          resulting from examination, appeal, litigation, or other sources.
                          Information from examinations of the largest corporations, which may
                          overlap information in the OTSA database, shows proposed income
                          adjustments in the tens of billions of dollars before reductions, but data
                          were not available from IRS on the results of examinations of smaller
                          corporations, partnerships, trusts, S corporations, or individuals.
                          Information from IRS’s contractor study estimates an annual tax gap due to
                          abusive shelters but has data and methodological limitations.

OTSA Database 	           As shown in table 1, as of September 30, 2003, an OTSA database included
                          estimated potential tax losses of about $33 billion from investments in
                          listed transactions, before considering any reductions resulting from
                          examination, appeal, litigation, or other sources and another $52 billion in
                          potential tax losses from nonlisted transactions with some characteristics
                          of abusive shelters. This database contains information on promoters and
                          investors and the amount of potential tax savings resulting from listed and
                          nonlisted transactions. Nonlisted transactions are transactions that
                          needed to be registered because they have some characteristics of abusive
                          shelters but were not, at least yet, determined to be abusive. According to
                          an IRS official, IRS was studying nonlisted transactions with about
                          $12 billion in potential tax losses for possible listing. The database only

                            For the decade from 1993 through 2002, corporations paid almost $2 trillion in income

                          Page 10                                                                       GAO-04-104T
                                          includes information on abusive or possibly abusive transactions that had
                                          been disclosed to or discovered by IRS.

Table 1: IRS’s Compilation of Tax Shelter Amounts as of January 14, 2003, and September 30, 2003

Category of transaction                       Number of transactions as of                       Potential tax loss (billions)a as of
                                          January 14, 2003        September 30, 2003           January 14, 2003b September 30, 2003
Listed                                                 3,423                        5,185                     $29.3                      $33.0
Nonlisted                                              1,334                        1,582                       44.7                      52.0
Total                                                  4,757                        6,767                     $73.9                      $85.0
Source: IRS OTSA database.
                                           The potential tax loss covers a multiyear period and does not consider reductions that may result from
                                          examination, appeal, litigation, or other sources.
                                          The numbers do not add to the total due to rounding.

                                          The estimated tax losses contained in the OTSA database cover a wide
                                          range of years from at least as far back as tax year 1989 and extending even
                                          to future tax years since, for instance, improperly claimed deductions may
                                          be used in some cases to reduce future taxes. For the $29 billion in
                                          estimated tax losses associated with listed transactions contained in the
                                          January 14, 2003, database, about 82 percent of the potential tax losses
                                          were concentrated in the period from 1993 through 2002.

                                          According to data IRS provided in mid-October 2003, OTSA had
                                          information on almost 300 firms that had possibly promoted abusive
                                          shelters as well as other tax planning products that contain at least some
                                          features of abusive transactions. It was also aware of about 6,400
                                          investors, including individuals and corporations that bought abusive
                                          shelters and other aggressive tax planning products.

Examinations of Large                     IRS has proposed shelter-related adjustments to large corporations’ income
Corporations                              in examinations it has closed and in examinations still open as of early May
                                          2003. In cases closed between October 1, 2001, and May 6, 2003, IRS
                                          proposed about $10.6 billion in abusive shelter-related adjustments to the
                                          income of 42 large corporations for tax years 1992-2000. These proposed
                                          adjustments would result in about $3.5 billion in tax revenue if the
                                          adjustments were not reduced. The corporations were in what is known as
                                          the Coordinated Industry Case (CIC) program, which includes the nation’s

                                          Page 11                                                                                 GAO-04-104T
                     largest corporations.12 They agreed with about $1.2 billion of the $10.6
                     billion in proposed adjustments to income.13 As of early August 2003,
                     Appeals research showed that few of the issues comprising the $9.4 billion
                     unagreed amount had been resolved yet by Appeals or through a settlement
                     initiative, although the database did not track all of them.14 For the 141
                     large corporations with cases still open in early May 2003, the amount of
                     proposed shelter-related income adjustments was $47.6 billion, translating
                     to about $16 billion in tax if not reduced. IRS did not have similar
                     information for smaller corporations. Also, since one of the sources of
                     information in the OTSA database is shelter-related adjustments proposed
                     in examinations, the proposed adjustments in the CIC program may
                     overlap the information in the OTSA database.

Contractor Study 	   In July 2003, an IRS contractor estimated the tax gap resulting from abusive
                     shelters for different years. For 1993 through 1999, based on the
                     contractor’s estimates, the average annual tax gap could have been as small
                     as about $11.6 billion or as large as about $15.1 billion of forgone tax.
                     However, the reliability of the contractor’s estimates is questionable
                     because of methodological and data constraints the contractor faced when
                     developing them.

                     The estimates followed a September 2001 recommendation by the Treasury
                     Inspector General for Tax Administration (TIGTA) that LMSB obtain a
                     more precise estimate of the shelter problem to lay a better foundation for
                     its strategy for addressing abusive shelters.15 In response, IRS contracted
                     for models to predict the likelihood of finding abusive shelters within

                      Under the CIC program, IRS continually audits about 1,100 of the nation’s largest
                     corporations, all of which have assets of more than $250 million.
                      IRS did not track the additional tax payments these corporations actually made related just
                     to the shelter-related adjustment. However, according to data provided by IRS, they paid
                     about an additional $552 million in taxes related to all issues raised by IRS, including the
                     abusive shelter issues.
                       In mid-August 2003, IRS gave us information showing that for the 14 abusive shelter
                     transactions Appeals had closed in fiscal year 2003 for CIC and other cases, Appeals
                     sustained about 71 percent of the dollar amounts proposed as adjustments to income.
                     Similar information was not available for earlier years.
                        TIGTA, Management Advisory Report: The Strategy for Curbing Abusive Corporate Tax
                     Shelter Growth Shows Promise but Could Be Enhanced by Performance Measures, Report
                     Number 2001-30-159 (Washington, D.C.: Sept. 13, 2001).

                     Page 12                                                                        GAO-04-104T
certain tax returns and to estimate the annual “tax gap” due to abusive
shelters. Both IRS and contractor officials believe the contract results are
more useful to predict returns with abusive shelters than they are to value
the size of the abusive shelter problem.

Nevertheless, as table 2 shows, the contractor produced estimates of the
size of the problem for each year from 1993 through 1999. Yearly low-end
estimates ranged from $9.0 billion of foregone tax in 1993 to $14.5 billion in
1999. On the other hand, the high-end estimates ranged from $12.1 billion
in 1993 to $18.4 billion in 1999.16 Averaging the estimates over time results
in the $11.6 billion to $15.1 billion range cited earlier.

Table 2: Contractor Estimates of the Size of the Abusive Shelter Problem (Dollars in

Year                                      Lower bound                             Upper bound
1993                                                $ 9.0                                  $12.1
1994                                                  9.5                                    12.7
1995                                                10,3                                     13.6
1996                                                11.4                                     14.9
1997                                                12.7                                     16.4
1998                                                13.6                                     17.3
1999                                                14.5                                     18.4
1993-1999 average                                   11.6                                     15.1
Source: Report provided by IRS.

Note: As computed by the contractor, the lower and upper bounds are the boundaries of 90 percent
confidence intervals associated with the estimates.

The tax gap model used three different kinds of data: (1) IRS’s Statistics of
Income data for the largest U.S. companies, those with assets over
$250 million falling within the CIC program, (2) Standard and Poor’s
Compustat financial data, and (3) surveys of IRS field offices. IRS
conducted surveys from 1999 through 2001 that asked field managers to
identify abusive tax shelters in their open inventory of examinations--

 Because the contractor found that estimating the problem’s size was difficult and
problematic, it applied a statistical technique to the estimates and produced other estimates
for each year. However, because it did not believe the statistical technique improved the
original estimates, we are not including the second set of estimates here.

Page 13                                                                            GAO-04-104T
relying on each manager’s understanding of what an abusive tax shelter is.
Since survey data are included in the OTSA database, some of the same
information used by the contractor appears in the OTSA information cited

Treasury, IRS, the contractor, and we have concerns about the contractor
estimates. First, it is difficult to determine whether these estimates might
be overstating or understating the true extent of the tax gap because of the
uncertainties in the underlying data and the elusive nature of the problem.
In identifying abusive shelters in the IRS surveys, field managers might
have anticipated that some abusive shelters existed where there were none
or where the assertion of abuse might not be sustained. On the other hand,
they might not have identified all the abusive shelters in their open
inventory of examinations because their definitions of abusive shelters
might have differed from each other. Finally, the data might not be
representative of all transactions, especially those that closed, because
survey responses were only to include open cases.

Second, the Statistics of Income data only included U.S. corporations with
assets of over $250 million falling within the CIC program. Many shelters
may be reflected in tax returns of smaller corporations, partnerships,
Subchapter S corporations, and wealthy individuals and were not included
in this study. Since these transactions were not included in the contractor’s
estimate, the resulting tax gap estimate is incomplete.

Third, the estimates are based on known shelters. They were developed
using 1990s’ ideas of what constituted abusive shelters. Since then, more
shelters have been disclosed or identified by IRS and still others are under
consideration for listing. Since the definition of an abusive shelter can
change over time, and the data cannot reflect unknown or unidentified
shelters, the operational definition of abusive shelters was a conservative

While the last two concerns argue that the contractor’s estimates
understate the true level of abusive shelters for recent years, the
contractor’s estimates and other indicators of the problem’s size based on
past data may also be of limited use as guides to current and future activity
for other reasons. According to Treasury and IRS officials, the legal and
economic environment has changed since the data for this study were
developed. First, they said, IRS has taken many administrative actions to
address abusive shelters. For instance, it is their belief that nothing puts
more of a damper on taxpayer participation in a particular type of

Page 14                                                           GAO-04-104T
                           transaction than IRS listing it. Similarly, although corporate-owned life
                           insurance transactions may heavily influence the contractor’s estimates,
                           legislation addressed the problem in 1996 and 1997, and therefore current
                           and future estimates would not reflect that problem—although they could
                           reflect problems not identified in the period covered by the contractor’s
                           study. Second, court cases have largely supported IRS’s assertions about
                           the need for business purpose requirements and about requirements for
                           economic substance in transactions. Third, today’s economy is not as
                           robust as the economy in the late 1990s, generating less profit to protect.
                           Finally, the publicity surrounding numerous corporate scandals may create
                           a chilling effect in the market for aggressive transactions. Countering these
                           points, however, are other opinions appearing in the press that (1) the
                           courts could uphold some tax shelters and (2) IRS’s capacity to stem
                           abusive shelters is limited.

IRS Strategy to Combat     IRS developed a broad-based strategy for combating abusive shelters that
                           included various features as well as elements of strategic planning.
Abusive Shelters Is        Deeming it a strategic initiative, IRS is executing a strategy incorporating
Broad-Based but            four principal elements: (1) an emphasis on promoters, (2) efforts to deter,
                           detect, and resolve abuse, (3) coordination of efforts throughout IRS, and
Generally Has No           (4) inducements provided for taxpayers to come forward and expedite case
Long-Term                  resolution. IRS is implementing a variety of initiatives designed to reduce
Performance Goals or       taxpayer incentives to participate in abusive transactions and discourage
                           promoters from marketing these transactions. Although IRS documents
Measures Linked to         outline an overall strategy for combating abusive shelters, IRS has
Goals                      generally not yet defined long-term performance goals for the effort and the
                           measures it would use to track progress in achieving those goals.17
                           However, IRS is planning to establish such goals and measures when it has
                           more information on the abusive shelter activities it is currently tracking.

IRS Is Actively Pursuing   IRS is actively pursuing abusive promoters to ensure (1) that tax strategies
Promoters                  containing characteristics of potentially abusive shelters are registered,
                           (2) that information about transactions is disclosed to IRS as required by

                              Although GPRA is generally applied to agencywide strategic plans, its framework is useful
                           to guide any type of planning. GPRA requires long-term strategic and annual performance
                           goals and associated measures, preferring measures relating to outcomes (results) versus
                           outputs (activities). The Office of Management and Budget says that strategic plans set out
                           long-term goals, outlining planned accomplishments and their implementation schedule.

                           Page 15                                                                        GAO-04-104T
                            sections 6111 and 6112 of the Internal Revenue Code, and (3) that,
                            according to IRS’s OTSA manager, those who generate noncompliance
                            change their behavior or go out of business. With 98 abusive shelter
                            promoters approved for investigation as of June 30, 2003, IRS uses
                            investigations to gain access to lists of the clients who buy promoters’
                            products and devise a roadmap to audit shelters included in the tax returns
                            of the investors. IRS is also using promoter investigations to enforce the
                            transaction registration requirements, which, in turn, assist in its efforts to
                            understand, track, and close abusive shelters. IRS announced the
                            completion of three large promoter investigations in 2001 through July
                            2003. They resulted in, among other things, three substantial payments and
                            promoter promises to work with IRS to ensure ongoing compliance with
                            shelter registration and list maintenance requirements.

IRS Efforts Are to Deter,   IRS focuses its efforts on deterring future marketing and sales of abusive
Detect, and Resolve         tax shelters and on detecting and resolving existing shelters. TIGTA
                            described IRS’s abusive shelter approach along the lines of deter, detect,
                            and resolve in September 2001.18 IRS considers its efforts to provide
                            guidance as early as possible to taxpayers and promoters in the form of
                            recently proliferating IRS and Treasury determinations, notices, and rulings
                            on abusive transactions and of registration, list maintenance, disclosure,
                            and other requirements to be a key deterrent. (See fig. 2.) Also designed to
                            deter abusive tax shelters, accuracy-related penalties aim at investors who
                            use abusive shelters to substantially undervalue true tax liability. Other
                            penalties are for promoters who market shelters that aid and abet the
                            understatement of tax liability or who fail to register shelters. IRS’s
                            Examination Returns Control System showed IRS assessing 21 investor
                            penalties totaling about $73 million between July 1, 2002, and May 1, 2003,
                            which taxpayers had not necessarily agreed to pay. During our review,
                            Treasury included proposed legislation in the Administration’s revenue
                            proposals to strengthen the penalties that could be used in abusive shelter

                            IRS’s ability to detect abusive shelters increased in the last 3 years due to
                            OTSA’s hotline, through which callers provide tips about transactions or
                            investors; disclosure, registration, and list maintenance requirements;
                            increased attention by IRS management; and increased use of IRS

                                 TIGTA, Report Number 2001-30-159.

                            Page 16                                                             GAO-04-104T
                          examination resources to look for shelter irregularities. For instance,
                          between May 31, 2000, and July 30, 2003, the hotline received 729 shelter-
                          related telephone calls and e-mails, some of them leading IRS to new listed
                          transactions, promoters, and investors. As another example, IRS expanded
                          its disclosure requirements in June 2002 to include noncorporate
                          taxpayers. Finally, as evidence of increased management attention, IRS
                          established a new senior position reporting to the IRS Chief Counsel to
                          supervise staff and lead task force initiatives to more quickly identify and
                          deal with abusive shelters.

                          Cases may be resolved at the examination level if taxpayers agree with IRS
                          findings. If taxpayers do not agree, cases are resolved at the appeals level,
                          through litigation, or by alternative dispute resolution.

                          In addition to these detection and case resolution efforts, IRS is using
                          Schedule K-1 data to research better methods of detecting abusive shelters
                          that involve multiple levels of flow-through entities.19 These complex
                          structures of related entities pose challenges in analyzing tax compliance
                          by creating opportunities for taxpayers to disguise noncompliance. In the
                          future, IRS hopes to use advanced data analysis tools such as link analysis
                          and graph-based data mining to identify potential abusive shelters. Link
                          analysis is the process of building networks of related entities, such as
                          flow-through entities and Schedule K-1 recipients, in order to expose
                          patterns and trends. Graph-based data mining, a form of link analysis, is
                          intended to enable IRS to identify structures of known abusive shelters and
                          find similar patterns in the population of flow-through networks to
                          discover previously undisclosed potential abusive shelter transactions. IRS
                          has paid a contractor $200,000 so far to assess the feasibility of these
                          technologies and plans to spend $575,000 over the next 1.5 to 2 years to
                          develop these concepts into models.

IRS Emphasizes Internal   Coordination within IRS and interface with Treasury on abusive shelters is
Coordination              a core objective in IRS’s plans for addressing those shelters. OTSA is the
                          focal point for all shelter-related activity performed in the Tax Shelter
                          Committee, the 6700 Committee, Counsel, Appeals, and LMSB. For
                          example, if a taxpayer discloses an investment in a tax shelter to IRS, OTSA
                          is to enter the transaction into its database, and OTSA reviews the

                           Appendix I describes the Schedule K-1, flow-through entities, and other compliance efforts
                          using Schedule K-1 data.

                          Page 17                                                                       GAO-04-104T
                             transaction in collaboration with IRS technical advisors and counsel.
                             OTSA may also forward it to LMSB examiners for compliance action.

                             At the IRS-wide level, an executive steering committee provides a forum
                             for coordinating work on both abusive shelters and abusive schemes. It
                             meets monthly and includes participants from LMSB, the Small
                             Business/Self Employed Division, Appeals, Counsel, and other
                             organizations. It operates under the auspices of IRS’s Enforcement
                             Committee, which was chartered in July 2003. Chaired by the Deputy
                             Commissioner for Services and Enforcement, a new position created in
                             May 2003, the Enforcement Committee is to guide IRS-wide enforcement
                             strategies, focusing on high-visibility issues involving many divisions or
                             potentially having significant compliance impact.

                             Although we did not systematically measure whether coordination is
                             facilitated by these mechanisms, we did review minutes of selected
                             executive steering committee meetings. In doing so, we saw such evidence
                             of coordination as the discussion of an LMSB and SB/SE working group on
                             who would work a corporate officer case when LMSB works on a

IRS Offers Inducements for   LMSB attempts to leverage its limited resources by using inducements to
Taxpayers to Disclose        achieve compliance. These tools include penalty relief, “fast track” issue
                             resolution, and various structured settlement programs that allow
Shelters and Expedite Case
                             participating taxpayers to keep a percentage of a shelter’s benefits in
Resolution                   exchange for conceding most benefits and expediting case resolution. For
                             example, under a disclosure initiative that expired on April 23, 2002,
                             taxpayers who revealed shelters and their respective promoters avoided
                             accuracy-related penalties. IRS’s aim was to more readily identify
                             promoters who had not registered shelters and, through the promoters,
                             find taxpayers who had not disclosed their shelter participation. As a
                             result of this initiative, IRS received 1,664 disclosures from 1,206 taxpayers,
                             disclosing tens of billions of dollars of losses and deductions.

                             IRS offered taxpayers various alternative dispute resolution mechanisms as
                             inducements to settle abusive shelter issues with IRS, mitigating the
                             hazards of litigation for both sides and moving more cases through the
                             administrative system quickly. For example, from October 2001 through
                             April 7, 2003, 17 taxpayers agreed with IRS on their respective shelter
                             issues in the Fast Track Issue Resolution program, resolving about
                             $1.6 billion in proposed adjustments to income (potentially about

                             Page 18                                                            GAO-04-104T
                           $540 million in tax). In another example, IRS announced initiatives in
                           October 2002 to resolve disputes related to three shelters: COLI, basis-
                           shifting shelters, and contingent liability shelters.20 In these initiatives, if
                           taxpayers agreed to settle their cases with IRS by a certain date, with the
                           last initiative closing March 5, 2003, they would pay a large percentage of
                           the full amount IRS disallowed. A summary as of early May 2003 of the
                           number of investors involved in the three settlement initiatives and the
                           potential tax dollars conceded or to be conceded appears in table 3.21

                           Table 3: Investors Accepting Abusive Shelter Settlement Initiative Offers and
                           Potential Tax Dollars Conceded or to Be Conceded as of Early May 2003

                                                                     Number of
                                                                      taxpayers    Number of taxpayers
                                                                      accepting       for whom IRS had
                                                                            IRS    information on taxes   Potential tax dollars
                                                                     settlement       conceded or to be     conceded or to be
                           Settlement initiative                           offer             conceded      conceded (billions)
                           COLI                                              24                     14                       $0.2
                           Basis shifting                                   267                     33                         0.6
                           Contingent liability                              62                     62                        2.8a
                           Total                                           353b                   109b                       $3.6
                           Source: Compiled by GAO from IRS data.

                           GAO estimated this number using an average of certain capital loss percentages to be conceded.

                           We do not know if a particular taxpayer was involved in more than one type of settlement initiative.

Generally IRS Does Not     Although IRS has outlined and begun to implement a multipart strategy for
Have Long-Term             combating tax shelters, it has not yet generally defined performance goals
                           for the effort and established the measures it would use to track progress in
Performance Goals or       achieving those goals. Performance goals define what an organization is
Measures Linked to Goals

                             IRS Notice 2001-51 identifies certain listed transactions. It describes basis-shifting
                           transactions as “certain redemptions of stock in transactions not subject to U.S. tax in
                           which the basis of the redeemed stock is purported to shift to a U.S. taxpayer.” It describes
                           contingent liability transactions as “transactions involving a loss on the sale of stock
                           acquired in a purported [Internal Revenue Code section] 351 transfer of a high basis asset to
                           a corporation and the corporation’s assumption of a liability that the transferor has not yet
                           taken into account for federal income tax purposes.”
                                Some of these investors are also included in the fast track program just described.

                           Page 19                                                                                 GAO-04-104T
trying to achieve over time, preferably focusing on the outcome desired
rather than activities or outputs. To date, according to IRS officials, their
shelter-related goals cover the number of staff years to be devoted to
shelter examinations and the number of shelter examinations to be closed.
Also, LMSB planning documents have a few short-term goals. For example,
LMSB had a short-term goal to begin compliance actions on all voluntary
shelter disclosures by June 30, 2003, a goal IRS officials told us was met.
IRS management officials recognize that developing other performance
goals and associated measures to track progress is desirable but point to
challenges they face in assessing the scope of the abusive shelter problem.
Nonetheless, IRS intends to establish such goals in the future when it has
more information on activities it is currently tracking.

IRS has already started down this road by developing several measures
that, while not tied to longer-term performance goals, are to be used in
tracking its progress in combating abusive tax shelters. It devised these
measures for fiscal year 2003 responding to a September 2001 TIGTA
recommendation to develop performance measures so managers could
better target problem areas, highlight successes, evaluate alternatives, and
track whether OTSA is achieving desired outcomes. IRS is mostly tracking
outputs related to case management, such as the number of tax shelter
examinations closed and tax shelter return cycle time, and is using output
measures of IRS program activities, such as published guidance issued and
hotline contacts. IRS is also using some measures that track tax
enforcement outcomes, namely adjustments proposed to tax returns from
disallowing abusive shelters and tax shelter penalties proposed.22 Since
fiscal year 2003 was the first year IRS used these measures, it had no
baseline data with which to evaluate its performance measures. However,
LMSB plans to evaluate its measures over time to assess their usefulness.

 LMSB called the tracking of adjustments a “record of tax enforcement results.” IRS does
not use performance measures for outcome measures like these because the IRS
Restructuring and Reform Act of 1998 prohibited it from using tax enforcement results to
evaluate any employee or to impose or suggest production quotas or goals.

Page 20                                                                      GAO-04-104T
Resource Shifts Are             Using admittedly limited information, IRS used a systematic decision-
                                making process in deciding to shift a large portion of LMSB examination
Significant but IRS             staff resources toward addressing abusive shelters. From fiscal year 2002
Faces Challenges in             through fiscal year 2004, LMSB expected to increase the portion of its
                                examination resources devoted to combating abusive shelters from 3
Addressing Abusive              percent in 2002 to 20 percent in 2004. In doing so, it will have shifted
Shelter Workload                resources out of examining the category of cases including such areas as
                                net operating losses and claims for refunds. Even so, IRS faces challenges,
                                especially in the near term, in addressing expected increases in its shelter
                                workload because of the growing number of shelter cases and limited
                                information it has on how long it takes to conduct shelter examinations. As
                                will be described, GAO has previously raised questions about IRS’s ability
                                to shift compliance resources as planned.

IRS Used Systematic             At an agencywide level, IRS decided staffing resource levels to be devoted
Planning and Budgeting          to addressing abusive shelters through a systematic planning and budgeting
                                process based on experience and professional judgment because IRS did
Process to Determine
                                not and does not have a reliable measure of the abusive shelter problem.
Staffing Priorities             Early in calendar year 2002, IRS’s divisions completed strategic
                                assessments in which they studied trends, issues, and priorities affecting
                                their operations. In April 2002, IRS’s senior management team, including
                                the Commissioner, Deputy Commissioner, division heads, and others used
                                two rounds of considering IRS’s programs to rank the needs for new or
                                redirected funding for fiscal year 2004. Of 33 programs considered, the
                                program including tax shelters received the third most votes. According to
                                an IRS official, this process also informed how funds already requested for
                                fiscal year 2003 would actually be spent. After the senior management
                                team reached consensus, the Commissioner issued overall planning
                                guidance for fiscal years 2003 and 2004 to reflect the jointly set strategic
                                direction, and the divisions wrote fiscal year 2003 and 2004 “strategy and
                                program plans” outlining staffing resources needed.

IRS Shifts Significant Levels   In 2002, LMSB put forward plans to increase its work on abusive shelters
of Examination Resources        from 3 percent of its examination resources to 20 percent between fiscal
                                years 2002 and 2004, assuming congressional funding. To support this shift
to Shelters                     in examination resources, LMSB needed to allocate examination resources
                                away from other areas. One area to receive less audit coverage was

                                Page 21                                                          GAO-04-104T
industry audits.23 As shown in table 4, from fiscal year 2003 to fiscal year
2004, IRS planned to move resources away from specific types of
mandatory examinations and from some high-risk nonmandatory returns.24
IRS’s strategy is to mitigate the impact of resource reallocations away from
nonshelter areas by using such issue management strategies as fast-track
resolution and prefiling agreements, thereby requiring less staff time to
close cases and freeing staff to be used in other areas.

Table 4: Percentage of LMSB Examination Resources in Different Examination

                                                                                               FY                FY      FY
Examination area                                                                             2002a             2003    2004
Shelters                                                                                         3%            15%     20%
Other mandatory examinations (including coordinated                                             N/A            55%     54%c
industry,b claims for refunds, net operating losses,
compliance initiative projects, and flow-through entities
related to wealthy individuals)
Related returns                                                                                 N/A             5%      4%c
High-risk, nonmandatory returns                                                                 N/A            15%     13%
Nonreturn examination activities                                                                N/A            10%     10%
Total                                                                                              --      100%       100%d
Source: LMSB September 20, 2002, presentation to the IRS Oversight Board, as amended after the presentation.
Information for most of the rows in this column was not available, as the presentation to the Oversight
Board did not include it.
 Coordinated industry cases are examinations of the nation’s largest corporations, those under
continual IRS audit.
  At the time of the September 20, 2002, presentation to the Oversight Board, the 54 and 4 percent
were 52 and 5 percent, respectively.
The column does not add to 100 percent because of rounding.

   IRS defines an “industry” case return as the return of an organization with assets of more
than $10 million but without being part of the largest corporations that are under continual
IRS audit.
 According to LMSB officials, mandatory examinations are those LMSB knows it will do,
such as those for abusive shelters and promoters. Nonmandatory examinations are what
remain after mandatory work is accommodated. High-risk nonmandatory examinations are
those in the nonmandatory category that have the highest probability that a taxpayer needs
compliance activity.

Page 22                                                                                                         GAO-04-104T
                          In addition to LMSB examination staff, IRS has managers, attorneys, and
                          others who work on abusive shelters. For instance, in February 2003,
                          OTSA and its parent body, the Office of Pre-Filing and Technical Guidance,
                          had 39 full-time and 34 part-time technical experts, program analysts, and
                          managers. Also at that time, a contact list for listed transactions included
                          17 attorneys. These numbers did not include many of the IRS legal
                          resources involved with abusive shelters. In addition, as of September 30,
                          2003, LMSB had assigned about 1,900 abusive and potentially abusive
                          shelter transactions involving non-LMSB taxpayers to IRS’s Small
                          Business/Self-Employed Division, which supplies examination staff
                          resources of its own.

IRS Faces Challenges in   Although IRS appeared to be on track to shift planned resources to shelter
Addressing Increasing     work in fiscal year 2003, it faces challenges in addressing the abusive
                          shelter workload, especially in the near term. This is because of (1) the
Shelter Workload
                          growing numbers of transactions and promoters to be examined and
                          (2) limited information on how long it takes to conduct shelter

                          From fiscal year 2002 through fiscal year 2004, LMSB planned to use 1,879
                          full-time equivalents (FTE) to address abusive shelters. During fiscal year
                          2002, LMSB used 239 FTEs to address tax returns that included abusive
                          shelters.25 According to IRS’s fiscal year 2004 congressional budget
                          justification, LMSB planned to allocate 691 and 949 FTEs in fiscal years
                          2003 and 2004, respectively. In a draft strategy and program plan dated
                          September 2003, LMSB projected it would actually use 615 FTEs for shelter
                          work in fiscal year 2003, or 88 percent of the planned amount and an
                          increase of 157 percent over the fiscal year 2002 FTE level including this

                          Because (1) the known abusive shelter workload has increased, (2) IRS has
                          limited experience to judge how many resources will be needed to work
                          the cases for how long a period, and (3) the workload may continue to
                          increase, it remains uncertain whether the substantial shift of resources to
                          shelter work will enable IRS to examine in a timely manner the growing
                          workload associated with shelters. For instance, the number of potential

                           According to LMSB officials, the fiscal year 2002 FTEs include time spent on the entire
                          returns containing shelters, not on the shelter issues alone. The estimates for fiscal years
                          2003 and 2004 are focused more on the shelter issues.

                          Page 23                                                                         GAO-04-104T
                                            examinations of listed transactions disclosed has grown since the inception
                                            of OTSA, adding significantly to IRS resources required to address the
                                            problem. Table 5 shows the number of listed transactions disclosed by
                                            taxpayers grew from 51 to 2,182 between December 31, 2000, and
                                            September 30, 2003, and other transactions disclosed to IRS grew from
                                            none to 663. The total of all listed and nonlisted LMSB-related transactions
                                            in the OTSA database, not only those disclosed by taxpayers, as of
                                            September 30, 2003, was 4,897.

Table 5: Taxpayer Disclosures of Listed and Other Reportable Transactions between 2000 and September 30, 2003

Section 6011 disclosures                  Calendar year (CY) 2000   CY 2001   CY 2002   CY 2003 through September 30   Total
Listed transactions disclosed                                 51         63     1,251                           817    2,182
Other reportable transactions disclosed                        0        214      308                            141     663
Source: IRS.

                                            IRS workload from promoter investigations has also grown since May 2002.
                                            At that time, IRS planned that 7 promoter investigations would be ongoing
                                            in fiscal year 2003. As of June 30, 2003, IRS had 98 promoter investigations
                                            approved. Based on early promoter investigations, an IRS official stated
                                            that promoter investigations can take thousands of hours to develop, and
                                            several have been litigated, each requiring a large expenditure of resources.

                                            LMSB has limited information on the amount of time required to examine
                                            abusive shelter cases. LMSB developed estimates of the amount of
                                            examination time required for such cases based on its experience
                                            examining various types of shelters but acknowledged that examiners can
                                            spend hundreds or thousands of hours depending on the type of shelter
                                            examined and the facts and circumstances of the case. For example,
                                            according to an LMSB official, based on personal experience, OTSA
                                            estimated that it would take about 800 hours to examine a potentially
                                            abusive transaction reflected in the return of a CIC corporation although
                                            LMSB had little data to support the estimate. During fiscal year 2003, IRS
                                            began collecting data on examination time that it plans to use for
                                            estimating the resources needed to address its abusive shelter workload.

                                            The future abusive shelter workload also could increase, at least in the
                                            short term. For example, as IRS learns more about the use of shelters, it
                                            may identify and list new kinds of transactions as being abusive. As IRS

                                            Page 24                                                             GAO-04-104T
                conducts the 98 promoter investigations approved as of June 2003, more
                investors are likely to be identified, and investor cases could lead to
                identifying more promoters. In addition, IRS expanded the types of
                taxpayers subject to disclosure requirements to include taxpayers like
                individuals, partnerships, and S corporations. According to IRS officials,
                disclosures from these types of taxpayers are first due to IRS for filing year
                2003 and generally do not yet appear in the OTSA database.

                In the longer term, what happens to the abusive shelter workload is less
                certain. To the extent that IRS actions and other factors reduce the size of
                the abusive shelter problem, IRS might not need to continue devoting as
                large a percentage of its examination resources to abusive shelters. How
                much and how soon such a drop may occur in abusive shelter cases is

                We have previously raised questions about IRS’s ability to shift compliance
                resources as planned. We recently testified that many parties have
                expressed concern about declining IRS compliance—especially audit—and
                collection trends for their potential to undermine taxpayers’ motivation to
                fulfill their tax obligations.26 Concerned about these trends, IRS has sought
                more resources, including increased staffing for compliance and
                collections since fiscal year 2001. Despite receiving requested budget
                increases, staffing levels in key occupations were lower in 2002 than in
                2000. These declines occurred for reasons such as unbudgeted expenses
                consuming budget increases and other operational workload increases.
                Based on past experience and uncertainty regarding some expected
                internal savings, fiscal year 2004 anticipated staff increases might not fully
                materialize. Thus, if IRS carries through with its intentions to increase
                resources devoted to abusive shelters, it may not have the desired level of
                resources in other areas of compliance.

Concluding 	    Abusive tax shelters represent a potentially significant, although
                imprecisely understood, loss in tax revenues. IRS developed and is
Observations	   following a broad-based, multifaceted strategy to combat abusive shelters
                even though it had limited data on the full scope of the problem. IRS’s
                strategy generally does not contain long-term performance goals and

                   U.S. General Accounting Office, Compliance and Collection: Challenges for IRS in
                Reversing Trends and Implementing New Initiatives, GAO-03-732T (Washington, D.C.:
                May 7, 2003).

                Page 25                                                                   GAO-04-104T
                    associated measures that can help Congress evaluate IRS’s progress.
                    Although establishing performance goals and measures is inherently
                    difficult since the scope and nature of abusive shelters is elusive, the need
                    for such goals and measures is heightened because IRS is shifting large
                    amounts of examination staff resources to support combating abusive
                    shelters. IRS’s initial decisions on shifting resources might need to be
                    reevaluated as IRS develops better information on the size of the abusive
                    shelter problem and the amount of time it takes to examine abusive shelter
                    cases. We encourage IRS to continue its efforts to obtain a better analytic
                    basis for determining the resources needed to address schemes and
                    shelters–while providing sufficient attention to other tax compliance
                    areas–and to develop goals and measures that it and Congress can use to
                    gauge IRS’s progress.

                    Mr. Chairman, this concludes my prepared statement. I would be happy to
                    respond to any questions you or other Members of the Committee may
                    have at this time.

Contact and 	       For further information on this testimony, please contact Michael Brostek
                    at (202) 512-9110 or brostekm@gao.gov. Individuals making key
Acknowledgements	   contributions to this testimony include Ralph Block, Elizabeth Fan, Amy
                    Friedheim, Lawrence Korb, Signora May, and James Ungvarsky.

                    Page 26                                                           GAO-04-104T
Appendix I

IRS Compliance and Research Programs
Using the Schedule K-1

              Schedule K-1s are information returns that link flow-through entities with
              their income recipients and therefore can be used for various compliance
              and research purposes, such as the automated underreporter (AUR)
              program1 and profiling potential nonfilers.

              Partnerships, S corporations, trusts, and estates are collectively known as
              flow-through entities because they can legally pass net income or loss
              through to their partners, shareholders, and beneficiaries. Flow-through
              entities are required to provide IRS and each partner, shareholder, or
              beneficiary with a Schedule K-1 stating the individual share of net income
              or loss to be reported. These individuals are then responsible for reporting
              this income or loss on their individual income tax returns and paying any
              applicable tax. According to IRS in tax year 2001, over 9 million flow-
              through entities reported passing through almost $1 trillion to
              approximately 24 million partners, shareholders, or beneficiaries. IRS
              research efforts suggest that 6 to 15 percent of the K-1s attached to flow-
              through returns are currently being omitted from beneficiary, partner, and
              shareholder returns. To better detect such noncompliance, IRS began
              transcribing nonelectronically submitted Schedule K-1s for tax year 2000 at
              a cost of about $20 million.

              In 2001, IRS added Schedule K-1 document matching to its AUR program.
              It began matching Schedule K-1 data to individual tax returns to identify
              taxpayers who had underreported flow-through income and had
              consequently underpaid their taxes. IRS estimated that K-1 matching
              program costs would be about $23.5 million total for both K-1 transcription
              and AUR program operations and that program yield would be $36 million
              in direct tax assessed. IRS also estimated that if voluntary compliance
              improved one percent due to the matching program, approximately
              $1.23 billion of additional tax would be generated annually. In the first year
              of the program, IRS issued about 69,000 notices to taxpayers and assessed
              about $29 million in additional taxes directly attributable to Schedule K-1
              underreporting.2 GAO estimates that when program assessments are
              compared to the costs of the program’s AUR operations, the return per

                The AUR program matches information return data, such as Forms W-2 and 1099 and
              Schedule K-1, with individual tax return data to verify that all income is reported.
               IRS began notifying taxpayers of potential discrepancies between income reported on the
              K-1 and individual tax returns in April 2002. However, after receiving complaints that
              notices were being sent to compliant taxpayers, IRS stopped issuing notices in August 2002.
              IRS data on number of notices sent and tax assessed were provided in August 2003.

              Page 27                                                                       GAO-04-104T
           Appendix I

           IRS Compliance and Research Programs 

           Using the Schedule K-1

           dollar of the K-1 matching program was about $9.31. If the cost of
           transcribing the K-1 data is included, the return per dollar decreases to
           about $1.25.3 Both of these assessment-to-cost ratios are substantially
           lower than that for the AUR program as a whole.4 The AUR program
           returned about $25 for every dollar spent in tax year 2000.5

           IRS has also used Schedule K-1 data to determine characteristics of
           potentially noncompliant taxpayer populations. Its preliminary profiling
           efforts identified over 227,000 business entities with almost $64 billion in
           Schedule K-1 income for tax year 2000 that potentially did not file tax
           returns. As of September 2003, IRS had begun to discuss ways of analyzing
           these cases to determine whether these businesses were required, but
           failed, to file returns, or whether inaccuracies in Schedule K-1 data
           produced false nonfiler leads. In addition, in response to a Treasury
           Inspector General for Tax Administration report issued in September 2002,6
           the agency has begun to research the effectiveness of using information
           returns, such as the K-1, to identify business nonfilers.

             To increase efficiency and improve the accuracy of K-1 data, IRS is exploring two-
           dimensional bar coding of Schedule K-1s. Instead of transcribing K-1 data, IRS would scan a
           bar code on the K-1 and electronically upload the information.
           Because the Schedule K-1 document matching program is new, its return on investment
           may be low compared to mature AUR programs.
            Information about the AUR program is based on IRS data from December 28, 2002.
           Treasury Inspector General for Tax Administration, The Internal Revenue Service Should
           Evaluate the Feasibility of Using Available Documents to Verify Information Reported on
           Business Tax Returns, Report Number 2002-30-185 (Washington, D.C.: Sept. 18, 2002).

(440158)   Page 28                                                                       GAO-04-104T
This is a work of the U.S. government and is not subject to copyright protection in the
United States. It may be reproduced and distributed in its entirety without further
permission from GAO. However, because this work may contain copyrighted images or
other material, permission from the copyright holder may be necessary if you wish to
reproduce this material separately.
GAO’s Mission	            The General Accounting Office, the audit, evaluation and investigative arm of
                          Congress, exists to support Congress in meeting its constitutional responsibilities
                          and to help improve the performance and accountability of the federal government
                          for the American people. GAO examines the use of public funds; evaluates federal
                          programs and policies; and provides analyses, recommendations, and other
                          assistance to help Congress make informed oversight, policy, and funding
                          decisions. GAO’s commitment to good government is reflected in its core values of
                          accountability, integrity, and reliability.

Obtaining Copies of       The fastest and easiest way to obtain copies of GAO documents at no cost is
                          through the Internet. GAO’s Web site (www.gao.gov) contains abstracts and full-
GAO Reports and           text files of current reports and testimony and an expanding archive of older
                          products. The Web site features a search engine to help you locate documents
Testimony                 using key words and phrases. You can print these documents in their entirety,
                          including charts and other graphics.
                          Each day, GAO issues a list of newly released reports, testimony, and
                          correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site
                          daily. The list contains links to the full-text document files. To have GAO e-mail this
                          list to you every afternoon, go to www.gao.gov and select “Subscribe to
                          e-mail alerts” under the “Order GAO Products” heading.

Order by Mail or Phone	   The first copy of each printed report is free. Additional copies are $2 each. A check
                          or money order should be made out to the Superintendent of Documents. GAO
                          also accepts VISA and Mastercard. Orders for 100 or more copies mailed to a single
                          address are discounted 25 percent. Orders should be sent to:
                          U.S. General Accounting Office
                          441 G Street NW, Room LM
                          Washington, D.C. 20548
                          To order by Phone: 	   Voice: (202) 512-6000
                                                 TDD: (202) 512-2537
                                                 Fax: (202) 512-6061

To Report Fraud, 	        Contact:
                          Web site: www.gao.gov/fraudnet/fraudnet.htm
Waste, and Abuse in       E-mail: fraudnet@gao.gov
Federal Programs          Automated answering system: (800) 424-5454 or (202) 512-7470

Public Affairs	           Jeff Nelligan, Managing Director, NelliganJ@gao.gov (202) 512-4800
                          U.S. General Accounting Office, 441 G Street NW, Room 7149
                          Washington, D.C. 20548
United States
                  Presorted Standard

General Accounting Office
      Postage & Fees Paid

Washington, D.C. 20548-0001

                                  Permit No. GI00

Official Business

Penalty for Private Use $300

Address Service Requested