Tax Compliance: Better Compliance Data and Long-term Goals Would 
Support a More Strategic IRS Approach to Reducing the Tax Gap	 
(18-JUL-05, GAO-05-753).					 
                                                                 
According to the Internal Revenue Service (IRS), a gap arises	 
each year between what taxpayers pay accurately and on time in	 
taxes and what they should pay under the law. The tax gap is	 
composed of underreporting of tax liabilities on tax returns,	 
underpaying of taxes due from filed returns, and nonfiling of	 
required tax returns altogether or on time. GAO was asked to	 
provide information on (1) the estimated amount that each major  
type of noncompliance contributed to the 2001 tax gap and IRS's  
views on the certainty of its tax gap estimates, (2) reasons why 
noncompliance occurs, and (3) IRS's approach to reducing the tax 
gap and whether the approach incorporates established		 
results-oriented planning principles.				 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-05-753 					        
    ACCNO:   A30179						        
  TITLE:     Tax Compliance: Better Compliance Data and Long-term     
Goals Would Support a More Strategic IRS Approach to Reducing the
Tax Gap 							 
     DATE:   07/18/2005 
  SUBJECT:   Data collection					 
	     Income taxes					 
	     Noncompliance					 
	     Personal income taxes				 
	     Policy evaluation					 
	     Strategic planning 				 
	     Tax administration 				 
	     Tax evasion					 
	     Tax nonpayment					 
	     Tax returns					 
	     Tax violations					 
	     Taxpayers						 
	     Voluntary compliance				 
	     IRS Taxpayer Compliance Measurement		 
	     Program						 
                                                                 
	     IRS National Research Program			 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-05-753

                 United States Government Accountability Office

              GAO	Report to the Committee on Finance, U.S. Senate

July 2005

TAX COMPLIANCE

 Better Compliance Data and Long-term Goals Would Support a More Strategic IRS
                        Approach to Reducing the Tax Gap

                                       a

GAO-05-753

[IMG]

July 2005

TAX COMPLIANCE

Better Compliance Data and Long-term Goals Would Support a More Strategic IRS
Approach to Reducing the Tax Gap

                                 What GAO Found

IRS estimates that underreporting of taxes accounted for about $250
billion to $292 billion of the $312 billion to $353 billion tax gap for
2001, while underpayment and nonfiling accounted for about $32 billion and
$30 billion, respectively. Although IRS has collected recent compliance
data, it still has concerns with some outdated methodologies and data used
to estimate the tax gap. IRS is taking laudable steps intended to improve
the estimate, which it plans to revise by the end of 2005. IRS has also
developed a proposed schedule of compliance studies, but it has no
approved plans to periodically measure compliance for the tax gap
components. While it may not be feasible or necessary to measure
compliance for all components at the same frequency or level of
investment, periodic compliance studies would support a more data-driven
and risk-based approach to reducing the tax gap.

IRS recently began to capture data on the reasons why taxpayers are
noncompliant. However, IRS has concerns about the data, such as examiners
assigning the same reason for noncompliance regardless of situation. Also,
it is often difficult for examiners to determine a taxpayer's
intent-whether the noncompliance is unintentional or intentional.
Collecting better data on reasons can help IRS focus its activities on
taxpayer service or enforcement. Although IRS is developing a system
intended to capture better examination data, IRS does not have firm or
specific plans to develop better reason data.

IRS approaches tax gap reduction through improving taxpayer service and
enforcing tax laws and has two broad strategic goals and related key
efforts that are intended to support this approach. However, IRS has not
established long-term, quantitative compliance goals and regularly
collected data to track its progress, which would complement its current,
important compliance efforts. Establishing clear goals and measuring
progress towards them would be consistent with results-oriented management
principles. IRS has begun to consider additional goals, but it is not yet
clear if they will be compliance related. Long-term, quantitative
compliance goals, coupled with updated compliance data, would provide a
solid base upon which to develop a more strategic, results-oriented
approach to reducing the tax gap.

IRS's Preliminary Tax Year 2001 Gross Tax Gap Estimates by Type of
Noncompliance and Type of Tax

Dollars in billions

                                       Type of tax                  
                  Type of Individual     Corporate                  
            noncompliance     income        income Estate    Excise     Total 
                                        Employment                  
           Underreporting $150-$187    $30 $66-$71     $4 No        $250-$292 
                                                          estimate  
             Underpayment         19           2 7      2         1       $32 
                                  28 No estimate        2 No              $30 
                Nonfiling            No estimate          estimate  

Total $198-$234 $32 $73-$78 $8 $1 $312-$353

Source: IRS.

Note: Figures may not sum to totals due to rounding.

United States Government Accountability Office

Contents

  Letter

Results in Brief
Background
Scope and Methodology
IRS Lacks Approved Plans to Regularly Measure Compliance,

Including Underreporting Which Accounts for the Largest
Portion of the Tax Gap
IRS Has Concerns with Its Data on Reasons for Noncompliance but
Does Not Have Firm or Specific Plans to Develop Better Data

IRS's Approach to Reducing the Tax Gap Focuses on Service and
Enforcement but Lacks Long-term, Quantitative Compliance
Goals and Measures That Are Consistent with Results-Oriented
Management Principles

Conclusions
Recommendations for Executive Action
Agency Comments and Our Evaluation

1

3 4 6

7

17

20 26 27 28

Appendix I IRS Compliance Measurement Surveys

Appendix II	Detailed Tax Gap Estimates, Data Sources, and Level of
Certainty

Appendix III IRS Key Efforts to Reduce the Tax Gap

Appendix IV Comments from the Internal Revenue Service

Appendix V GAO Contact and Staff Acknowledgments 39

  Tables

Table 1: IRS's Preliminary Tax Year 2001 Gross Tax Gap Estimates
by Type of Noncompliance and Type of Tax 7

Table 2: Data Sources for IRS's Preliminary Tax Year 2001 Tax Gap
Estimates by Type of Noncompliance and Type of Tax, and IRS's Level of
Certainty in the Estimates 10

Table 3: Types of Surveys by Return Type and Year 30

Table 4: IRS's Preliminary Tax Year 2001 Gross Tax Gap Estimates, Data
Sources, and Level of Certainty by Tax Gap Component and Type of Tax 32

Abbreviations

DCE detection controlled estimation
EITC Earned Income Tax Credit
GAO Government Accountability Office
GPRA Government Performance and Results Act of 1993
IRS Internal Revenue Service
NRP National Research Program
SOI Statistics of Income
TCMP Taxpayer Compliance Measurement Program
TY Tax Year

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.

United States Government Accountability Office Washington, DC 20548

July 18, 2005

The Honorable Charles Grassley
Chairman
The Honorable Max Baucus
Ranking Minority Member
Committee on Finance
United States Senate

The federal tax system relies on taxpayers to voluntarily comply with the
tax laws. However, a gap arises each year between what taxpayers pay
accurately and on time in taxes and what they should pay under the law.
Recognizing the need for current compliance data to update the tax gap
estimate, the Internal Revenue Service (IRS) implemented a new
compliance study in 2002 called the National Research Program (NRP) to
produce such data for tax year 2001 while reducing taxpayer burden.1 NRP
is a significant achievement and its data should be valuable in improving
IRS operations and for other uses.

Incorporating preliminary results from NRP, IRS recently estimated a
"gross" tax gap from $312 billion to $353 billion for tax year 2001.2 IRS
estimated that it would eventually recover some of this amount through
late payments and IRS enforcement actions, resulting in an estimated "net"
tax gap for 2001 from $257 billion to $298 billion.3 The tax gap estimate
is
an aggregate of estimates for the three primary types of noncompliance:4
(1) underreporting of tax liabilities on tax returns; (2) underpaying of
taxes due from filed returns; and (3) nonfiling, which refers to the
failure
to file a required tax return altogether or on time.

1GAO, Tax Administration: New Compliance Research Effort Is on Track, but
Important Work Remains, GAO-02-769 (Washington, D.C.: June 27, 2002).

2IRS's most recent estimates of the tax gap are preliminary, and as such,
IRS presents them as ranges.

3Throughout this report, references to the tax gap refer to the gross tax
gap unless otherwise noted.

4Estimates for each type of noncompliance include estimates for some or
all of the five types of taxes that IRS administers-individual income,
corporate income, employment, estate, and excise taxes. Throughout this
report, references to the tax gap estimate refer to the aggregate
estimate, unless otherwise noted.

The tax gap arises when taxpayers fail to comply with the tax laws, either
intentionally or unintentionally. As a result of their noncompliance, the
burden of funding the nation's commitments, including funding growing
budget deficits, falls more heavily on taxpayers who voluntarily pay their
taxes. In addition, IRS expends substantial resources enforcing and
explaining tax laws, with the goals of increasing compliance and reducing
the tax gap.

Given your concern over the burden that the tax gap places on both the
taxpayers who voluntarily pay their taxes and the federal budget, we
testified before the Senate Committee on Finance on April 14, 2005, on a
number of tax-gap-related issues.5 As you requested, this report
elaborates on the testimony by providing additional information on (1) the
estimated amount that each major type of noncompliance contributed to the
2001 tax gap and IRS's views on the certainty of its tax gap estimates,
(2) reasons why the noncompliance occurs, and (3) IRS's approach to
reducing the tax gap and whether the approach incorporates established
results-oriented planning principles.

To provide information on the estimated amount that each type of
noncompliance contributed to the 2001 tax gap, we reviewed IRS's tax gap
estimates for 2001. To determine IRS's views on the certainty of its tax
gap estimates, we reviewed IRS studies and interviewed IRS research
officials about tax gap estimation. We reviewed IRS, academic, and our
prior work and interviewed IRS officials in an attempt to identify the
various reasons for taxpayer noncompliance. To determine IRS's approach to
reducing the tax gap and whether the approach incorporates established
resultsoriented planning principles, we examined IRS's strategic and
performance plans and interviewed IRS officials. We asked IRS to identify
its key efforts to reduce the tax gap as well as the related rationales,
goals, and results. Using what we learned about IRS's approach, we
determined the extent to which the approach incorporated selected planning
principles consistent with the Government Performance and Results Act of
1993 (GPRA).6 We conducted our review from June 2004 through May 2005 in
accordance with generally accepted government auditing standards.

5GAO, Tax Compliance: Reducing the Tax Gap Can Contribute to Fiscal
Sustainability but Will Require a Variety of Strategies, GAO-05-527T
(Washington, D.C.: Apr. 14, 2005).

6Pub. L. No. 103-62 (1993).

  Results in Brief

For the 2001 gross tax gap estimate of about $312 billion to $353 billion,
IRS estimated in March 2005 that underreporting accounted for about $250
billion to $292 billion while underpayment and nonfiling accounted for
about $32 billion and $30 billion, respectively. The actual tax gap could
be higher or lower due to various factors that affect IRS's certainty of
the estimate. For example, due to a lack of reliable data, IRS's estimate
does not include some types of noncompliance, such as corporate income tax
nonfiling. Also, IRS is concerned with some of the outdated data and
methodologies used to estimate the tax gap. Finally, it is difficult for
IRS to identify and measure noncompliance, such as underreported income,
when IRS has little or no information from third parties about payments
made or taxes withheld. IRS is taking some steps, such as updating the
data and methodology for estimating individual income tax underreporting,
that IRS intends to use to revise the preliminary tax gap estimate during
2005. While IRS has proposed a schedule for NRP studies over the next
several years, it has no approved plans to regularly measure tax
compliance, the results of which it could use to update the tax gap
estimate, identify new or growing areas of noncompliance, and make
informed decisions about resource allocations to address noncompliance.

IRS has concerns about its data on the reasons why taxpayers do not comply
with tax laws. Taxpayer noncompliance can be unintentional or intentional
in various ways. For example, taxpayers might unintentionally err on their
tax returns because they misunderstand the laws or guidance explaining
compliance requirements, or improperly omit income from their returns
based on poor advice from tax practitioners. Alternatively, taxpayers may
intentionally omit income from their tax returns to evade taxes. IRS
captures data on the various unintentional and intentional reasons for
noncompliance during examinations of tax returns. IRS is concerned with
the reliability of the data since, for example, some examiners have
assigned the same reason for all noncompliance, regardless of the
situation. Also, determining taxpayer intent-whether the noncompliance is
unintentional or intentional-can be difficult. Although IRS is developing
a system intended to capture better examination data, IRS has no firm or
specific plans to develop better data on the reasons why taxpayers do not
comply, through steps such as improved data entry controls and examiner
training. Without such data, it is more difficult for IRS to decide
whether its efforts to address specific areas of noncompliance should
focus on nonenforcement activities, such as improved forms or
publications, or enforcement activities.

IRS's approach to reducing the tax gap includes improving taxpayer service
to increase voluntary compliance and enhancing enforcement of

tax laws by detecting and addressing noncompliance, but does not
incorporate some steps consistent with results-oriented management. To
support this approach, IRS has established two broad strategic goals and
identified over 40 related key efforts, which include using direct
enforcement actions to address high-income nonfilers and using analytical
models to pursue higher priority collection cases. However, IRS has not
established long-term, quantitative compliance goals and regularly
collected data to track progress in reducing the tax gap, which would
complement its current, important compliance efforts. Establishing
longterm, quantitative compliance goals and measuring progress towards
them offer several benefits to both IRS and external stakeholders and
would be consistent with the performance management principles set forth
in GPRA. Although IRS faces challenges in implementing a results-oriented
management approach to reducing the tax gap, IRS's recently collected
compliance data provide an improved foundation for setting compliance
goals and reexamining programs intended to reduce the tax gap.

We are making recommendations that IRS develop plans to periodically
measure tax compliance, take steps to improve its data on the reasons why
taxpayers do not comply, and establish long-term, quantitative goals for
voluntary compliance levels with a focus on individual income tax
underreporting and total tax underpayment. Taken together, these steps can
help IRS build a foundation to understand how its taxpayer service and
enforcement efforts affect compliance, improve the efforts, and make
progress on reducing the tax gap. The Commissioner of Interval Revenue
agreed with our recommendations, highlighted challenges associated with
them, and commented on various steps IRS would take to implement each
recommendation.

Background 	IRS develops its tax gap estimate by measuring the rate of
taxpayer compliance-the degree to which taxpayers fully complied with
their tax obligations. IRS uses such compliance data, along with other
data and assumptions, to estimate the dollar amount of taxes not paid
accurately and on time. For tax year 2001, IRS estimated that from 83.4
percent to 85 percent of owed taxes were paid voluntarily and on time, and
that from $312 billion to $353 billion in taxes were not paid that should
have been. IRS also estimates the amount of the gross tax gap that it will
recover through enforcement and other actions and subtracts that to
estimate the net annual tax gap. For tax year 2001, IRS estimated that it
would

eventually recover about $55 billion for a net tax gap from $257 billion
to $298 billion. As we have reported in the past,7 closing the entire gap
may not be feasible since it could entail more intrusive recordkeeping or
reporting than the public is willing to accept or more resources than IRS
is able to commit. However, given the size of the tax gap, even modest
reductions would yield very significant financial benefits.

IRS has estimated the tax gap on multiple occasions, beginning in 1979.
IRS's earlier tax gap estimates relied on the Taxpayer Compliance
Measurement Program (TCMP), through which IRS periodically performed
line-by-line examinations of randomly selected tax returns. TCMP started
with tax year 1963 and examined individual returns most frequently-
generally every 3 years-through tax year 1988. IRS contacted all taxpayers
selected for TCMP studies. IRS did not implement any TCMP studies after
1988 because of concerns about costs and burdens on taxpayers.8

Under NRP, a program that we have encouraged, IRS recently completed its
initial review of about 46,000 randomly selected individual tax returns
from tax year 2001 (see app. I for a list of conducted TCMP and NRP
surveys). Unlike with TCMP studies, IRS did not need to contact taxpayers
for every tax return selected under NRP, handled some taxpayer contacts
through correspondence rather than face-to-face examinations, and
generally only asked taxpayers to explain information that it was
otherwise unable to verify through IRS and third-party databases. In
addition, unlike operational examinations, NRP examinations were randomly
selected and used to measure compliance rather than target suspected
noncompliance.

IRS has a strategic planning process through which it supports decisions
about strategic goals, program development, and resource allocation. Under
GPRA,9 agencies are to develop strategic plans as the foundation for
results-oriented management. GPRA requires that agency strategic plans
identify long-term goals, outline strategies to achieve the goals, and
describe how program evaluations were used to establish or revise the

7GAO, Taxpayer Compliance: Analyzing the Nature of the Income Tax Gap,
GAO/T-GGD-97-35 (Washington, D.C.: Jan. 9, 1997).

8GAO, Tax Administration: Status of IRS' Efforts to Develop Measures of
Voluntary Compliance, GAO-01-535 (Washington, D.C.: June 18, 2001).

9Pub. L. No. 103-62 (1993).

  Scope and Methodology

goals. GPRA requires federal agencies to establish measures to determine
the results of their activities.

To provide information on the estimated amount that each major type of
noncompliance contributed to the 2001 tax gap, we reviewed IRS's tax gap
estimate for 2001. To determine IRS's views on the certainty of its
estimate, we reviewed IRS studies about tax gap estimation and interviewed
IRS research officials to understand the data and methodologies used. We
also spoke with IRS officials regarding planned changes to the data
sources and estimation methodologies for the tax gap estimate. We
determined that the tax gap estimates presented in this report are
sufficiently reliable for the specific purposes of our engagement,
particularly since IRS already has publicly released its tax gap estimates
and disclosed their weaknesses. These purposes include discussing the
major tax gap components and the orders of magnitude for various
components, IRS's concerns about the certainty of its estimates, and our
recommendations on IRS's compliance data and efforts.

We reviewed IRS, academic, and our own reports and interviewed IRS
officials to identify the various reasons for noncompliance. We talked
with IRS officials to determine the extent and reliability of data and
coding on the reasons for noncompliance, and reviewed IRS's Examination
Operational Automation Database, which is a database of tax return
examination results that includes examiners' determinations of the reasons
for any noncompliance. We also talked with IRS officials to determine any
plans to develop better data on reasons for noncompliance.

To determine IRS's approach to reducing the tax gap and whether the
approach incorporates established results-oriented planning principles, we
reviewed IRS strategic and performance plans and interviewed IRS strategic
planning officials at the agency and operating division levels. We asked
IRS to identify its key efforts to reduce the tax gap as well as the
related rationales, goals, and results. As part of our work on whether the
approach incorporates established results-oriented planning principles, we
used what we learned about IRS's approach to determine the extent to which
it incorporated selected planning principles consistent with GPRA's
requirements. For purposes of this review, we focused on elements of
results-oriented planning that, previously, we found common to leading
organizations successfully pursuing results-oriented management-defining

  IRS Lacks Approved Plans to Regularly Measure Compliance, Including
  Underreporting Which Accounts for the Largest Portion of the Tax Gap

desired results, measuring performance, and using performance information
to support agency missions.10

IRS estimates that underreporting of taxes accounted for about $250
billion to $292 billion of the $312 billion to $353 billion tax gap for
2001, while underpayment and nonfiling accounted for about $32 billion and
$30 billion, respectively. The actual tax gap could be higher or lower due
to various factors that affect the certainty of the estimate, such as old
compliance data. IRS is taking some steps designed to improve portions of
its compliance measurement efforts and its preliminary tax gap estimate
and plans to release a revised tax gap estimate by the end of 2005. While
IRS has proposed a schedule for NRP studies over the next several years,
IRS has no approved plans to regularly measure tax compliance, which it
could use to update the tax gap estimate and guide its compliance efforts.

Underreporting Accounted As table 1 indicates, underreporting of
individual income taxes
for Most of the Tax Gap represented about half of the tax gap for 2001
(the estimate ranges from
Estimate $150 billion to $187 billion out of a gross tax gap estimate that
ranges from

                         $312 billion to $353 billion).

Table 1: IRS's Preliminary Tax Year 2001 Gross Tax Gap Estimates by Type
of Noncompliance and Type of Tax

                              Dollars in billions

                                  Type of tax

      Type of     Individual  Corporate                             
noncompliance  income tax income tax Employment  Estate  Excise      Total 
                                        tax         tax      tax    
Underreporting  $150-$187        $30     $66-$71     $4    No    $250-$292 
                                                           estimate 
    Underpayment          19          2           7      2        1       $32 
     Nonfiling            28         No No estimate      2    No          $30 
                               estimate                    estimate 
       Total       $198-$234        $32     $73-$78     $8       $1 $312-$353 

Source: IRS.

Note: Figures may not sum to totals due to rounding.

10GAO, Executive Guide: Effectively Implementing the Government
Performance and Results Act, GAO/GGD-96-118 (Washington, D.C.: June 1996).

Within the underreporting estimate, IRS attributed about $150 billion to
$187 billion, or about 50 percent of the total tax gap, to individual
income tax underreporting, including underreporting of business income,
such as sole proprietor,11 informal supplier,12 and farm income (about $83
billion to $99 billion); nonbusiness income, such as wages, interest and
capital gains (about $42 billion to $57 billion); overstated income
adjustments, deductions, and exemptions (about $14 billion to $16
billion); and overstated credits (about $11 billion to $14 billion).
Underreporting of corporate income tax contributed an estimated $30
billion, or about 10 percent, to the 2001 tax gap, which included both
small corporations (those reporting assets of $10 million or less) and
large corporations (those reporting assets of over $10 million). (For a
more detailed table of IRS's estimates for the various components of the
2001 tax gap, see app. II).

Employment tax underreporting accounted for an estimated $66 billion to
$71 billion, or about 20 percent, of the 2001 tax gap and included several
taxes that must be paid by self-employed individuals and employers.
Selfemployed individuals are generally required to calculate and remit
Social Security and Medicare taxes to the U.S. Treasury each quarter.
Employers are required to withhold these taxes from their employees'
wages, match these amounts, and remit withholdings to Treasury at least
quarterly. Underreported self-employment13 and employer-withheld
employment taxes respectively contributed an estimated $51 billion to $56
billion and $14 billion to IRS's tax gap estimate. The employment tax
underreporting estimate also includes underreporting of federal
unemployment taxes (about $1 billion).

11Sole proprietors are self-employed individuals who should file a
Schedule C with their individual tax return to report profits and losses
from their business. Sole proprietors include those who provide services,
such as doctors or accountants; produce goods, such as manufacturers; and
sell goods at fixed locations, such as car dealers and grocers.

12Informal suppliers are sole proprietors who work alone or with few
workers and, by definition, operate in an "informal" manner. Informal
suppliers include those who make home repairs, provide child care, or sell
goods at roadside stands. These taxpayers should report business profits
or losses on a Schedule C.

13As employment taxes and income taxes for self-employed taxpayers are
largely assessed on the same income, self-employed individuals who
underreport their income consequently underreport the employment tax due
on that income.

IRS's 2001 Tax Gap Although a significant portion of IRS's new tax gap
estimate is based on Estimate Is Inexact Due to recent compliance data,
IRS has concerns with the certainty of the overall Incomplete and Old
Data, tax gap estimate in part because of incomplete and old data,
outdated Outdated Methodologies, methodologies, and measurement
difficulties. Table 2 shows IRS's and Inherent Measurement certainty level
in the estimates, as well as the underlying data sources.14

Difficulties

14 For a discussion of the data sources IRS used to estimate the tax gap,
see Internal Revenue Service, Understanding the Tax Gap, FS-2005-14 (March
2004), http://www.irs.gov/newsroom/article/0,,id=137246,00.html
(downloaded Mar. 30, 2005); Internal Revenue Service, Federal Tax
Compliance Research: Individual Income Tax Gap Estimates for 1985, 1988,
and 1992, Publication 1415 (Rev. 4-96) (Washington, D.C.: Apr. 1996); and
Robert E. Brown and Mark J. Mazur, IRS's Comprehensive Approach to
Compliance Measurement (Washington, D.C.: June 2003),
http://www.irs.gov/pub/irssoi/mazur.pdf (downloaded June 6, 2005).

Table 2: Data Sources for IRS's Preliminary Tax Year 2001 Tax Gap
Estimates by Type of Noncompliance and Type of Tax, and IRS's Level of
Certainty in the Estimates

Type of noncompliance and IRS certainty type of tax Estimate data
source(s) level

Underreporting -

                                       c

Individual income tax  o  Tax Year (TY) 2001 NRP Survey

o  TY 1988 and earlier TCMP studies

o  1981, 1985-6 University of Michigan surveys on informal suppliers

o  1984 University of Illinois study on tip income

                                       -

                                       c

Corporate income tax o    TY 1977 and 1980 TCMP surveys (only for   Weaker 
                                       small corporations)             
                        o      Operational examinations (only for      
                               mid-sized and large corporations) -     
                             averaged over 3 years in the mid-1980s    
                        o   TY 1982 TCMP study of unrelated business   
                                    income tax of tax-exempt           
                                          organizations                
                        o       TY 1975 TCMP study on fiduciaries      

Employment tax  o  TY 2001 NRP Survey

o  TY 1984 withholding noncompliance study

o  1981 and 1985-6 University of Michigan surveys on informal suppliers

o  1984 University of Illinois study on tip income

                                       -

                                       c

Estate tax  o  	IRS's Statistics of Income (SOI) program data for filed
estate tax returns for Reasonable TY 1992

            Excise tax            o        No estimate         Not applicable 
    Underpayment (all types of    o      IRS Master File       Actual figures 
              tax)a                                           
            Nonfilingb                                             Reasonable 
      Individual income tax       o  TY 1988 Nonfiler (TCMP)       Reasonable 
       Corporate income tax       o        No estimate         Not applicable 
          Employment tax          o        No estimate         Not applicable 

Estate tax  o  	2 University of Michigan longitudinal surveys (begun in
1992 and 1993 and Reasonable interviews participants every 2 years)

o  TY 1992 IRS's SOI

                   Excise tax  o  No estimate Not applicable

Source: IRS.

a

Unlike the other components of the 2001 tax gap, the underpayment
component is not an estimate, but rather represents the tax amounts that
taxpayers reported on time, but did not pay on time.

b IRS's nonfiler estimate for individual income tax is net of amounts of
true tax liability that are paid on time (e.g., through withholding).
Refunds that are due to nonfilers do not reduce the nonfiling gap, since
they are not associated with a tax liability.

c

These estimates are based on more recent NRP data, but IRS has not
finalized the certainty level for these estimates because it has not yet
completed its assessment of the quality of the NRP data.

Tax Gap Estimate Is Incomplete

Estimates for Some Components of the Tax Gap Are Based on Old Data

As table 2 shows, IRS's estimate for the 2001 tax gap does not include
estimates of excise tax underreporting and nonfiling. According to IRS,
the reason for this omission is that numerous federal excise taxes, many
of which have specific exclusions or varying applications, complicate
excise tax computations. Further, data on excise tax transactions are
typically maintained at the state level and are often incomplete. Also,
according to an IRS research official, the estimate does not include
corporate income tax and employment tax nonfiling because IRS lacks good,
representative data for corporate and employment tax nonfilers. Further,
data from IRS's operational programs to identify nonfilers exclude those
whom IRS does not know about and do not include the full tax liability of
nonfilers whom IRS has identified.

The 2001 tax gap estimate also does not include any estimates for taxes
due from illegal source income, as the magnitude of such income is
difficult to estimate.15 Moreover, the federal government seeks to
eliminate most illegal activities altogether, rather than derive revenue
from these

16

activities.

Old data also contribute to IRS's "weaker" level of certainty for certain
segments of the underreporting portion of its 2001 tax gap estimate. For
example, IRS used data from the 1970s and 1980s to estimate underreporting
of corporate income taxes and employer-withheld employment taxes. For
large corporate income tax underreporting, IRS based its estimate on the
amount of tax recommended from operational examinations rather than the
tax ultimately assessed as part of the total tax liability. According to
IRS officials, IRS relies on the amount of tax recommended because it is
difficult to determine the true tax liability of large corporations due to
complex and ambiguous tax laws that create opportunities for differing
interpretations and that complicate the determination. These officials
further stated that because these examinations are not randomly selected
and are not focused on identifying all tax noncompliance, the estimate
produced from the examination data is not representative of the tax gap
for all large corporations. They also explained that due to these
complexities and the costs and burdens of collecting complete and accurate
data, IRS has not systematically measured large corporation tax compliance
through statistically valid

15Illegal source income may include drugs, illegal gambling, prostitution,
etc.

16IRS's Criminal Investigation division pursues illegal activities that
have tax consequences, but does not measure the revenue generated by the
cases it pursues.

Tax Gap Estimates Are Affected by Outdated Methodologies

Tax Gap Is Inherently Difficult to Estimate

studies, even though the officials acknowledged that such studies would be
useful in estimating the related tax gap.17

Further, some methodologies IRS used to estimate the tax gap are based on
older data and contribute to the uncertainty surrounding the tax gap
estimate. For example, because IRS knew that it would not detect all
underreporting noncompliance, IRS multiplied the detected amounts of
underreporting to help calculate a total estimate for underreported
individual income tax. IRS officials explained that they used a number of
"multipliers," including one derived from the 1976 TCMP study of
individual tax returns, which was before IRS expanded and improved its
computer matching programs to better detect various types of underreported
income.18 In addition, IRS estimated individual income tax nonfiling based
on the assumption that the relationship between individual income
nonfiling and underreporting has been constant since the 1988 TCMP survey
was conducted.

Finally, it is inherently difficult for IRS to observe and measure some
types of underreporting or nonfiling. For example, underreporting of
income or nonfiling of tax returns by informal suppliers can be hard for
IRS to detect because the tax laws generally do not require third parties
to withhold income tax or file information returns on payments made to
informal suppliers, as are required with other types of individuals such
as wage earners. Similarly, academic studies have discussed the difficulty
in tracking cash payments that businesses make to their employees, as
businesses may not report these payments to IRS in order to avoid paying
employment taxes and employees may not report these payments on their
income tax return to avoid paying income taxes.

17GAO, Tax Administration: Compliance Measures and Audits of Large
Corporations Need Improvement, GAO/GGD-94-70 (Washington, D.C.: Sept. 1,
1994); Tax Administration: Factors Affecting Results from Audits of Large
Corporations, GAO/GGD-97-62 (Washington, D.C.: Apr. 17, 1997); Tax
Administration: IRS Measures Could Provide a More Balanced Picture of
Audit Results and Costs, GAO/GGD-98-128 (Washington, D.C.: June 23, 1998).

18IRS's computer matching programs use third-party information documents
to verify information reported on tax returns. IRS established the
multiplier by comparing the amount of income detected through TCMP
examinations conducted without information documents and matching the
income detected with the aid of these tools.

    IRS Plans to Issue a Revised Tax Gap Estimate but Has No Approved Plans to
    Regularly Collect Compliance Data

IRS is taking several steps that could improve the preliminary tax gap
estimate for tax year 2001. IRS intends to publish a revised tax gap
estimate by the end of 2005 based on the results of these steps.

For example, IRS officials stated that IRS plans to further analyze the
preliminary NRP results in an attempt to improve the certainty of the
estimate. NRP is a significant achievement and its data should be valuable
in improving IRS operations and for other uses. However, those officials
added that because IRS is still assessing the quality of the NRP data, it
has not yet finalized the certainty levels for the preliminary estimates
for individual income tax and self-employment tax underreporting.
Likewise, we cannot yet be certain about the quality of the NRP data
collected because IRS is still assessing the data.

IRS plans to implement three changes to its estimation methodology for its
revised tax gap estimate. Although it is too soon to know whether these
changes will improve the estimate, IRS expects that the changes will help
address known methodological weaknesses. According to IRS, these changes
include the following:

o  	IRS plans to replace the multiplier it derived in the 1970s and used
to estimate individual income tax underreporting. IRS is developing a new
methodology, known as detection controlled estimation (DCE). DCE is a
regression-based model that will use 2001 NRP data and control for
variables that could affect the amount of underreporting detected.19

o  	IRS plans to develop a new technique as well as replace the data from
the 1981 and 1985-1986 University of Michigan surveys to estimate the
individual income tax underreporting portion of the tax gap attributable
to informal suppliers.

o  	IRS intends to update its estimate of individual income tax nonfiling,
which is currently based on 1988 nonfiler TCMP data, by using "Exact
Match" data provided by the U.S. Census Bureau.20 Census will match data
from its Current Population Survey against the IRS Master Files to
identify the extent of nonfiling by individual taxpayers. The Census data
to be provided to IRS will be aggregated and not contain information on
specific individuals.

19 By fall 2005, IRS plans to have determined which variables to include
in the DCE model.

20IRS has used "Exact Match" data for past tax gap estimates.

In addition, IRS research officials are planning a compliance measurement
study that will allow IRS to update underreporting estimates involving
flow-through entities. This study, which IRS intends to begin in October
2005, would take 2 to 3 years to complete. Because individual taxpayers or
corporations may be recipients of income (or losses) from flow-through
entities, this study could affect IRS's underreporting estimates for
individual and corporate income tax.

While these data and methodology updates could improve the tax gap
estimates, IRS has no approved plans to periodically collect more and
better compliance data over the long term beyond the study of flowthrough
entities. IRS Research officials said that they recently proposed a
schedule for additional NRP studies over the next several years. However,
these officials also said this proposal is under consideration but has not
been finalized. IRS has indicated that given its current research
priorities, it could not begin another NRP study of individual income tax
returns before 2008, at the earliest, and would not complete such a study
until at least 2010.

According to IRS officials, IRS has not committed to regularly collecting
compliance data because of the associated costs and burdens. Taxpayers
whose returns are examined through compliance studies such as NRP bear
costs in terms of time and money. Also, IRS incurs costs, including direct
costs and opportunity costs (or revenue that IRS potentially forgoes by
examining randomly selected returns, which are more likely to include
returns from compliant taxpayers than returns selected because they are
likely to contain noncompliance that would produce additional tax
assessments).

    Regularly Measuring Compliance Can Be Beneficial, but Determining the
    Frequency of Measurement Involves Considering Several Factors

Regularly measuring compliance can offer many benefits, including helping
IRS identify new or growing types of noncompliance, identify changes in
tax laws and regulations that may improve compliance, more effectively
target examinations of tax returns, understand the effectiveness of its
programs to promote and enforce compliance, and determine its resource
needs and allocations.21 For example, by analyzing 1979 and 1982 TCMP
data, IRS identified significant noncompliance with the number of
dependents claimed on tax returns and justified a legislative change to
address the noncompliance. As a result, for tax year 1987,

21GAO, Tax Administration: IRS' Plans to Measure Tax Compliance Can Be
Improved, GAO/GGD-93-52 (Washington, D.C.: Apr. 5, 1993).

taxpayers claimed about 5 million fewer dependents on their returns than
would have been expected without the change in law.

Tax compliance data are useful outside of IRS as well. Other federal
agencies and offices use compliance data for tax policy analysis, revenue
estimating, and research. For example, the Department of Commerce's Bureau
of Economic Analysis had used TCMP data to adjust its national income and
product accounts.22 Additionally, state tax authorities have used IRS
compliance data to develop state compliance programs and estimate state
tax gaps. Also, policy makers in the executive branch and Congress can use
the results from compliance measurement studies to help decide on
appropriate funding levels for IRS.

As we have reported in the past, the longer the time between compliance
measurement surveys, the less useful they become given changes in the
economy and tax law.23 According to IRS, without current compliance data,
it has limited capability to determine key areas of noncompliance to
address and actions to take to maximize the use of its limited resources.
For example, the formulas that IRS creates from compliance data to select
returns for examination have enabled IRS to focus examination resources on
noncompliant returns rather than burdening compliant taxpayers. When IRS
updated the formulas in the early 1990s with compliance data from the 1988
TCMP, IRS selected a lower percentage of compliant tax returns for
examination. However, after 3 years of using formulas based on the 1988
data, the percentage of compliant tax returns examined increased each year
through 1998, placing additional burdens on compliant taxpayers and
leaving less time for IRS to examine noncompliant returns that resulted in
an additional tax assessment.

Historically, IRS has varied how frequently it measured compliance for
particular types of taxpayers and taxes. As appendix I shows, the period
between measurements of individual income tax reporting compliance, which
consistently has accounted for the largest portion of the tax gap, never
exceeded 4 years between 1963 and 1988. In planning the 2001 NRP to
measure individual income tax compliance, IRS envisioned doing the NRP on
a 3-year cycle. Appendix I also shows that IRS measured compliance less
frequently for other types of taxpayers and taxes, such as

22 These accounts include measures of personal income that are used to
allocate funds for a number of federal programs.

23GAO/GGD-93-52.

for small corporation income taxes, and that IRS never measured compliance
for large corporations or for excise taxes.

Although regularly measuring tax compliance can be beneficial, how often
measurements should be made is a judgment that depends on many potential
criteria including (1) the amount that a particular type of noncompliance
is thought to contribute to the tax gap, (2) whether IRS has reason to
believe that compliance may have changed (e.g., due to tax law changes),
and (3) costs, particularly when IRS officials said that resources to
conduct operational examinations are already limited. Using these
criteria, IRS would likely vary the frequency of compliance measurement
studies. Based on these criteria as well as our previous reports,24
decisions about compliance measurement would also be affected by the
following factors.

o  	Precision. The costs and benefits of measuring compliance can vary
with how precisely IRS wishes to measure compliance to achieve an intended
use (e.g., tax gap estimation or examination return selection). Obtaining
more precise and more detailed compliance data for more detailed
populations of taxpayers or tax issues (e.g., types of income or
deductions) would likely be more costly but potentially more useful.

o  	Capacity. Each compliance measurement study requires having enough
resources such as staffing, training, tools, and systems to capture the
data. Regular compliance measurement through smaller efforts targeted at
particular types of taxpayers or taxes and sampling designs that collect
data across consecutive tax years rather than for one year could help
reduce costs and sustain long-term compliance measurement.

24GAO/GGD-93-52; GAO, Tax Compliance: Status of the Tax Year 1994
Compliance Measurement Program, GAO/GGD-95-39 (Washington, D.C.: Dec. 30,
1994); and GAO, Tax Administration: Alternative Strategies to Obtain
Compliance Data, GAO/GGD-96-89 (Washington, D.C.: Apr. 26, 1996).

  IRS Has Concerns with Its Data on Reasons for Noncompliance but Does Not Have
  Firm or Specific Plans to Develop Better Data

Several factors concern IRS about its data on the reasons for
noncompliance, which can be unintentional or intentional. Although IRS is
developing a system intended to capture better examination data, IRS does
not have firm or specific plans to develop better data on the reasons for
noncompliance, even though the lack of such data makes it harder to decide
whether it should address specific areas of noncompliance through
nonenforcement efforts, such as designing clearer forms or publications,
or enforcement efforts.

    IRS Has Concerns with Its Data on Reasons for Noncompliance

IRS has concerns with its data on the unintentional and intentional
reasons for noncompliance. Various types of unintentional or intentional
reasons could explain why taxpayers fail to comply with the tax laws.25
Unintentional reasons can include being unaware of recordkeeping
requirements, accidentally entering an item on the wrong line of a tax
return, or following inaccurate advice from a tax practitioner.
Intentional reasons for noncompliance can include intentionally omitting
income from a tax return or interpreting vague tax laws to evade tax
liability.

IRS collects data on the reasons for noncompliance for specific tax issues
during its operational examinations of tax returns.26 In many of these
cases, it is difficult for examiners to determine a taxpayer's intent-
whether the noncompliance is unintentional or intentional. Unless the
evidence clearly points to the reason, the examiner would have to make
subjective judgments about why the noncompliance occurred. IRS has a
number of other concerns with the data:

25Academic research on the reasons for taxpayer noncompliance is fairly
limited. That research includes a "typology of noncompliance," developed
by Robert Kidder and Craig McEwen, to describe the various categories of
noncompliance. These categories include procedural (failure to follow
rules on which forms to file); taxpayer laziness; classic tax cheating;
brokered (involves use of a tax preparer); symbolic (due to perceived
unfairness in the tax laws); and social (based on the extent that
taxpayers believe others are complying with the law). See Robert Kidder
and Craig McEwen, "Taxpaying Behavior in Social Context: A Tentative
Typology of Tax Compliance and Noncompliance," in Jeffrey A. Roth and John
T. Scholz, Eds. Taxpayer Compliance, Volume 2: Social Science Perspectives
(Philadelphia, Pa.: University of Pennsylvania Press, 1989).

26 IRS also collected reason data in NRP, but we did not determine to what
extent IRS's concerns about the reason data from operational examinations
also applied to NRP data.

o  	The database is incomplete because not all examination results,
including data on reasons for noncompliance, were being entered into the

27

database.

o  	IRS has not tested the adequacy of the controls for data entry or the
reliability of the data being collected. IRS has found instances where
examiners close examinations without assigning a reason for noncompliance
or by assigning the same reason to all instances of noncompliance,
regardless of the situation.28

o  	IRS has not trained all examiners to ensure consistent understanding
and use of the various codes to indicate the reason for noncompliance.

o  	The data do not represent the population of noncompliant taxpayers but
rather only those who had their tax returns examined.

    IRS Does Not Have Firm or Specific Plans to Develop Better Data on Reasons
    for Noncompliance, which Could Help IRS's Efforts to Address Tax
    Noncompliance

According to IRS officials, the agency does not have firm or specific
plans to develop better data on the reasons for noncompliance. One
official explained that IRS decided not to improve the consistency of its
current reason data because it is devoting its limited resources to other
efforts, such as developing the Examination Desktop Support System (EDSS).
Although this system is intended to allow examiners to capture better
examination data, specific system features have not yet been identified to
improve examiners' selection of reason codes. IRS officials said that the
system could be enhanced in the future to improve the reason data and that
they plan to consider such enhancements.

As the National Taxpayer Advocate recently testified,29 data on whether
taxpayers are unintentionally or intentionally noncompliant with specific
tax provisions are critical to IRS for deciding whether its efforts to
address specific areas of noncompliance should focus on nonenforcement
activities, such as improved forms or publications, or enforcement

27 In October 2004, IRS started implementing a system to improve case
processing and data capture, particularly for adjusted tax amounts.

28 An IRS official said that managers are to review the accuracy of the
data entry of examination results but that they do not know the extent to
which managers actually review the entry of reason data.

29 Testimony of Nina E. Olson, National Taxpayer Advocate, before the
Senate Committee on Finance, April 14, 2005.

activities to pursue intentional noncompliance. For example, taxpayers may
unintentionally claim the Earned Income Tax Credit (EITC) because they do
not understand the child residency requirements for this credit (i.e., a
dependent must live with the taxpayer for more than half of the year).
This type of unintentional noncompliance may require IRS to more clearly
explain the EITC requirements within related forms and publications.
However, other taxpayers may file false EITC claims with the intent of
evading tax liability, which may suggest a strategy that relies on IRS's
enforcement programs and tools. Similar situations could exist for other
tax code provisions.

If IRS is to develop better data on the reasons for noncompliance, it will
be important for IRS to consider factors in data collection such as the
following.

o  	Data reliability. To minimize examiner subjectivity and ensure that
the data are complete and accurate, IRS would need to refine the reason
categories, provide adequate training, establish system and data entry
controls, and provide supervisory oversight.

o  	Scope. IRS would need to decide whether the reason categories are to
be captured for selected types of noncompliance or all types of
noncompliance.

o  	Examination selection. IRS currently collects reason data annually
through hundreds of thousands of operational examinations. IRS also
collected reason data through NRP. In the future, IRS would need to decide
whether to collect reason data (1) during all operational examinations,
(2) for a statistical sample of operational examinations, or (3) for
examinations performed through periodic compliance studies such as NRP.
Collecting data for a sample of examinations or through periodic
compliance studies might be done with a smaller cadre of examiners
specially trained and overseen to maximize consistency of decisions about
the reasons why taxpayers are noncompliant. Also, data from samples of
examinations could be used to generalize reasons for noncompliance for all
examinations, and data from compliance studies of all taxpayers could be
used to generalize these reasons for the population of taxpayers.

Our past reports30 have supported the concept of rigorously researching
the causes of noncompliance. Recognizing the benefits of better

30GAO, Tax Research: IRS Has Made Progress but Major Challenges Remain,
GAO/GGD-96-109 (Washington, D.C.: June 5, 1996).

  IRS's Approach to Reducing the Tax Gap Focuses on Service and Enforcement but
  Lacks Long-term, Quantitative Compliance Goals and Measures That Are
  Consistent with Results-Oriented Management Principles

compliance data, the National Taxpayer Advocate has also urged IRS to
consider performing additional research into causes of noncompliance.31

IRS approaches tax gap reduction through improving service to taxpayers
and enforcing tax laws and has established two broad strategic goals and
related key efforts that are intended to support this approach. However,
IRS has not established long-term, quantitative compliance goals and
regularly collected data to track progress in reducing the tax gap, which
would complement its current important compliance efforts. Establishing
clear compliance goals and measuring progress towards them benefits both
IRS and external stakeholders and are consistent with the resultsoriented
performance management principles set forth in GPRA. Although IRS has
lacked such data in the past and faces other challenges, NRP and EITC data
provide an improved base for setting compliance goals and reexamining
existing programs intended to reduce the tax gap.

    IRS Approaches Tax Gap Reduction through Broad Goals and Numerous Efforts to
    Improve Taxpayer Service and Enforce Tax Laws

IRS's overall approach to reducing the tax gap consists of improving
service to taxpayers and enhancing enforcement of the tax laws. Through
efforts such as education and outreach programs, IRS seeks to improve
voluntary compliance with the tax system by helping people understand
their tax obligations. In addition, IRS attempts to simplify the tax
process, such as by revising forms and publications to make them more
easily understood by diverse taxpayer communities and electronically
accessible. In conjunction with taxpayer service, IRS uses its enforcement
authority to ensure that taxpayers are reporting and paying the proper
amount of taxes. Through efforts such as examining tax returns and
collaborating with state governments to share leads on abusive tax
avoidance transactions, IRS seeks to detect and deter noncompliance.

31Testimony of Nina E. Olson, National Taxpayer Advocate, before the
Senate Committee on Finance, July 21, 2004, and Internal Revenue Service,
Taxpayer Advocate Service, National Taxpayer Advocate 2004 Annual Report
to Congress (Washington, D.C.: Dec. 31, 2004).

Two of IRS's three strategic goals, along with their associated objectives
and strategies, are intended to directly support this approach.32

o  	Goal 1-Improve Taxpayer Service-is intended to promote voluntary
compliance. This goal has three objectives (1) improve service options for
the tax paying public (2) facilitate participation in the tax system by
all sectors of the public and (3) simplify the tax process.

o  	Goal 2-Enhance Enforcement of the Tax Law-is intended to ensure,
through IRS's enforcement authority, that taxpayers are meeting their tax
obligations. The four objectives for this goal are (1) discourage and
deter noncompliance with emphasis on corrosive activity by corporations,
highincome individual taxpayers, and other contributors to the tax gap;
(2) ensure that attorneys, accountants, and other tax practitioners adhere
to professional standards and follow the law; (3) detect and deter
domestic and off-shore-based tax and financial criminal activity; and (4)
deter abuse within tax-exempt and governmental entities and misuse of such
entities by third parties for tax avoidance or other unintended purposes.
To achieve these objectives, IRS has 15 strategies, such as "re-examine
and adjust audit processes to target likely areas of noncompliance."

In addition to these goals, IRS's service and enforcement efforts outlined
in its strategic plan are also intended to support tax gap reduction.
IRS's strategic plan mentions over 60 service and enforcement efforts
targeted at improving taxpayer compliance. Because the plan did not
prioritize these efforts, we asked IRS officials to identify the key
efforts in reducing the tax gap. In response, IRS provided over 40 key
efforts. Enforcement efforts included pursuing high income nonfilers
(taxpayers with income over $100,000 who have not filed a tax return)
through direct enforcement actions and identifying higher priority
collection cases through analytical models. Service, or nonenforcement,
efforts included a taxpayer education program on tip reporting. (See app.
III for a summary of the key efforts provided.)

32 Modernization objectives and strategies under Strategic Goal 3 are
intended to support tax gap reduction by helping IRS manage its employee
and technology resources effectively and efficiently. Because this goal
helps IRS meet its service and enforcement goals, this report does not
discuss the goal separately.

IRS's Tax Gap Reduction IRS has developed a strategic planning and
budgeting process33 to help the Approach Does Not agency comply with GPRA
requirements. However, IRS's strategies for Establish Long-term, improving
compliance generally lack a clear focus on long-term,

quantitative goals and results measurement. IRS has established
broadQuantitative Compliance qualitative goals and strategies for
improving taxpayer service andGoals or Regular Data enhancing enforcement
of the tax laws. IRS has also identified measures,Collection to Measure
such as compliance rates for tax reporting, filing, and payment as well as
Progress the percentage of Americans who think it is acceptable to cheat
on their taxes,34 which are intended to gauge the progress of its
strategies toward its broad goals. However, IRS does not collect recent
data to update all of these compliance measures and has not established
quantitative goals against which to compare the measures and judge any
progress made through its compliance strategies.

Although IRS has not focused on quantitative, results-oriented goals for
improving voluntary compliance, IRS has established many output-related
goals and measures that track activity level, such as the number of
taxpayers contacted, collection cases closed, or returns examined. In
contrast, IRS has fewer outcome-related goals and measures that track
results, such as refund timeliness or examination quality.

In the past, IRS had set a long-term goal of improving overall compliance
to 90 percent by 2001. This goal was to be achieved through a research
approach rooted in IRS's Compliance 2000 philosophy.35 The Compliance 2000
philosophy envisioned using nonenforcement efforts to correct
unintentional noncompliance and reserving enforcement efforts for

33IRS implemented a new strategic planning, budgeting, and performance
management process during fiscal year 2000. The process begins with the
operating divisions preparing strategic assessments. After receipt and
review of the strategic assessments, the commissioner provides detailed
guidance to the operating divisions for developing their strategy and
program plans. These plans are then incorporated into an IRS-wide
performance plan (which sets out measurable objectives such as the number
of audits to be done). These plans are, in turn, incorporated into IRS's
budget justification (which sets out its resource requests to Congress).
The remaining steps involve allocating resources across IRS divisions and
programs and monitoring division adherence to the planning and budgeting
decisions.

34Other measures for enhancing enforcement are (1) the percentage of
priority guidance list items published (percentage of tax issues IRS will
address through regulations, notices, and other forms of guidance) and (2)
average cycle time between receipt and completion of an audit case.

35GAO/GGD-96-109; GAO, Tax Administration: Compliance 2000-A Worthy Idea
that Needs Effective Implementation, GAO/T-GGD-92-48 (Washington, D.C.:
June 3, 1992).

intentional noncompliance. To carry out this philosophy, in the early
1990s, IRS initiated many research projects across IRS's 63 district
offices to identify noncompliant market segments, root causes for the
noncompliance, and innovative ways to improve compliance. However, the
lack of objective compliance data, among other factors, limited the
success of this approach. Recently, external stakeholders, such as the IRS
Oversight Board, have supported the concept of setting a numeric, longterm
goal for increasing the voluntary compliance rate.

In response to a President's Management Agenda36 initiative to better
integrate budget and performance information, IRS officials said that they
are considering various long-term goals for the agency. IRS has not yet
set a time frame for publicly releasing the goals.37 Nor have IRS
officials indicated whether any goals will be related to improving
taxpayer compliance or whether they will be quantitative and
results-oriented.

    Long-term, Quantitative Compliance Goals and Measures Are Beneficial and
    Consistent with Results-Oriented Management Principles

Focusing on outcome-oriented goals and establishing measures to assess the
actual results, effects, or impact of a program or activity compared to
its intended purpose can help agencies improve performance and
stakeholders determine whether programs have produced desired results. As
such, long-term, quantitative compliance goals offer several benefits for
IRS, as discussed below.

Perhaps most important, compliance goals coupled with periodic
measurements of compliance levels would provide IRS with a better basis
for determining to what extent its various service and enforcement efforts
contribute to compliance. Additionally, long-term, quantitative goals may
help IRS consider new strategies to improve compliance, especially since
these strategies could take several years to implement. For example, IRS's
progress toward the goal of having 80 percent of all individual tax
returns electronically filed by 200738 has required enhancement of its
technology,

36The President's Management Agenda is intended to help the federal
government become more results-oriented and encourage federal managers to
ask whether their programs are working as intended and, if not, what can
be done to achieve greater results.

37 According to IRS officials, developing long-term, results-oriented
goals is a complex process that requires sustained management commitment.
These factors contribute to IRS's uncertainty about when it will publicly
release its goals.

38Congress established this electronic filing goal in the IRS
Restructuring and Reform Act of 1998, Pub. L. No. 105-206 (1998).

development of software to support electronic filing, education of
taxpayers and practitioners, and other steps that could not be completed
in a short time frame. Focusing on intended results can also promote
strategic and disciplined management decisions that are more likely to be
effective because managers who use fact-based performance analysis are
better able to target areas most in need of improvement and select
appropriate interventions. Likewise, agency accountability can be enhanced
when both agency management and external stakeholders such as Congress can
assess an agency's progress toward meeting its goals. Finally, setting
long-term, quantitative goals would be consistent with results-oriented
management principles that are associated with highperforming
organizations and incorporated into the statutory management framework
Congress has adopted through GPRA.

    Updated Compliance Data Provide Opportunities for IRS to Establish a More
    Results-Oriented Tax Gap Reduction Approach

Not unlike other agencies we have reported on in the past,39 IRS faces
challenges in implementing a results-oriented management approach, such as
identifying and collecting the necessary data to make informed judgments
about what goals to set and to subsequently measure its progress in
reaching its goals. However, having completed the NRP review of income
underreporting by individuals, IRS now has an improved foundation for
setting goals for improving taxpayers' compliance.40

IRS's effort to address noncompliance with the EITC provides an example of
how a more data-driven planning approach can help IRS become more
results-oriented over time.41 IRS's most recent EITC compliance study
estimated that between $8.5 billion and $9.9 billion, or between 27
percent and 32 percent, respectively, of the EITC claims filed for tax
year 1999 should not have been paid. Following the release of this study,
a task force

39GAO, Results-Oriented Government: GPRA Has Established a Solid
Foundation for Achieving Greater Results, GAO-04-38 (Washington, D.C.:
Mar. 10, 2004).

40The Internal Revenue Service Restructuring and Reform Act of 1998, Pub.
L. No. 105-206 (1998), specifically prohibits IRS from using its records
of tax enforcement results to evaluate employees or to impose or suggest
production quotas or goals with respect to such employees. That
restriction does not, however, prevent IRS from using its records of tax
enforcement results to examine whether its current enforcement efforts are
effective in deterring noncompliance and to in turn establish long-term
strategies and priorities for improvement.

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

of IRS and Treasury officials determined the three leading types of errors
that accounted for about $7 billion annually in overclaims. On the basis
of compliance data and other research, IRS started an initiative to
improve service, fairness, and compliance and designed specific corrective
actions targeting the three types of errors. IRS is evaluating these
actions to determine their effectiveness at reducing the overclaim rate in
each of the three errors. Because IRS targeted its EITC effort based on
data on the sources and extent of taxpayer errors, it was better able to
determine what actions to take and how well, using systematic data
collection and program evaluation, the effort is meeting its intended
purpose.

Measuring progress toward any goals that may be set could be challenging.
For example, IRS researchers have found it difficult to determine the
extent to which its enforcement actions deter noncompliance or its
services improve compliance among taxpayers who want to comply. Although
widespread agreement exists that IRS enforcement programs generally
increase voluntary tax compliance, challenges such as collecting reliable
compliance data, developing reasonable assumptions about taxpayer
behavior, and accounting for factors outside of IRS's actions that can
affect taxpayer compliance, such as changes in tax law, make it difficult
to estimate the effect of IRS's enforcement and service activities. Even
if IRS is unable to empirically estimate the extent to which its actions
directly affected compliance rates, periodic measurements of compliance
levels can indicate the extent to which compliance is improving or
declining and provide a basis for reexamining existing programs and
triggering corrective actions if necessary.

Recently, several research studies have offered insights to better
understand the direct tax revenue effects of IRS's activities as well as
the indirect effects on voluntary tax compliance.42 IRS researchers have

42 Two types of indirect effect are (1) the increase in voluntary
compliance in the larger population resulting from examinations, or other
enforcement and nonenforcement actions, on targeted taxpayers, and (2) the
increase in voluntary compliance of the targeted taxpayer in subsequent
years. Economists have estimated the indirect effect of an examination on
voluntary compliance to range between 6 and 12 times the amount of the
proposed adjustment. See Alan H. Plumley, The Determinants of Individual
Income Tax Compliance: Estimating The Impacts of Tax Policy, Enforcement,
and IRS Responsiveness, Publication 1916 (Rev. 11-96), (Washington, D.C.:
Nov. 1996), 2, 35-36; Jeffrey A. Dubin, Michael J. Graetz and Louis L.
Wilde, "The Effect of Audit Rates on the Federal Individual Income Tax,
1977-1986," 43 National Tax Journal, (1990), 395, 396, 405; and Jeffrey A.
Dubin, "Criminal Investigation Enforcement Activities and Taxpayer
Noncompliance" (paper written for the IRS Research Conference, June 2004),
http://www.irs.gov/pub/irs-soi/04dubin.pdf (downloaded July 1, 2005).

Conclusions

hypothesized that the indirect effect of an examination varies among
taxpayer segments. Further, a recent study concluded that criminal
investigations have positive direct and indirect tax effects. Although
these studies generally indicate that IRS activities have positive tax
effects, the magnitude of these effects is not yet known with a high level
of confidence given compliance measurement challenges, as mentioned
earlier. According to IRS, these studies serve as a valuable baseline for
further research, but it has not yet determined how it will use these
studies to make operational decisions.

As discussed in our recent testimony on the tax gap before the Senate
Committee on Finance, and underscored by IRS, periodic tax compliance
measurement is critically important to IRS's ability to estimate the tax
gap and design compliance programs intended to reduce the tax gap. Without
current, reliable compliance data, it can be difficult for IRS to monitor
trends or identify new types of noncompliance, determine its compliance
resource needs and how to allocate such resources, and justify budget and
staffing requests to policy makers in Congress and the executive branch.
Consequently, completion of NRP, which covered the largest portion of the
tax gap and was designed and implemented with an eye to reducing the costs
and burdens of data collection, is a substantial achievement. However,
although IRS has recently proposed a schedule for future NRP studies, it
has no approved plans to repeat this study or periodically measure
compliance across the various components of the tax gap. Doing periodic
compliance studies in areas that have previously been measured, such as
individual income tax underreporting, would provide valuable information
to support a more data-driven and risk-based approach towards improving
compliance and reducing the tax gap. Although it may not be feasible or
necessary to measure compliance for all components of the tax gap at the
same frequency or with the same level of investment, where practical
methodologies exist, periodic measurements should be taken. Where
practical methodologies do not yet exist, such as for excise tax or for
large corporations, looking for ways to overcome challenging compliance
measurement difficulties would be worthwhile.

The tax gap is both a measure of the burden and frustration of taxpayers
who want to comply but are tripped by tax code complexity and of willful
tax cheating by a minority who do not wish to pay their fair share to
support government programs. As such, collecting data on the reasons why
noncompliance occurs can help IRS more effectively tailor its efforts to
improve compliance. It can be difficult for IRS examiners to consistently
determine the reasons why taxpayers have failed to comply

with the tax laws. However, IRS has no specific plans to address this
issue and, as a result, is missing opportunities to gather better data
than it already collects. Certain immediate steps, like improving reason
codes, better training examiners in applying the codes, and possibly
reducing the number of examiners who would be responsible for making
judgments on the reasons taxpayers are noncompliant may improve the data
IRS currently collects. Nevertheless, given the difficulty of consistently
determining why taxpayers are noncompliant, sustained research on these
reasons also would be needed to develop a better understanding.

Reducing the tax gap will be a challenging task given persistent levels of
noncompliance and will not likely be achieved through a single solution.
Rather, the tax gap must be attacked on multiple fronts and with multiple
strategies over a sustained period of time. Without long-term,
quantitative voluntary compliance goals and related performance measures,
it will be more difficult for IRS to determine the success of its
strategies, adjust its approach when necessary, and remain focused on
results, especially since factors that affect compliance change over time.
Having compliance goals, coupled with recently collected NRP data, would
provide a solid base upon which IRS can develop a more strategic,
results-oriented approach to reducing the tax gap.

Taken together, these steps-periodically measuring compliance, determining
the reason taxpayers are noncompliant, and setting resultsoriented
long-term goals-can help IRS build a foundation to understand how its
taxpayer service and enforcement efforts affect compliance, improve its
efforts, and make progress on reducing the tax gap.

Recommendations for To establish a stronger foundation for improving IRS's
efforts to reduce the tax gap, the Commissioner of Internal Revenue should
do the Executive Action following.

o  	Develop plans to periodically measure tax compliance for areas that
have been previously measured, such as for individual income tax
underreporting, and study ways to cost effectively measure compliance for
other components of the tax gap that have not been measured, such as for
excise tax and large corporations. Those plans and that study should take
into account risk management factors such as the amount the component
contributes to the gap, changes that may have affected compliance levels
since a measurement was last taken, and the cost of measuring compliance.

o  	Take steps to ensure that IRS regularly collects complete, accurate,
and consistent data, to the extent possible, on the reasons taxpayers are
noncompliant and that sufficient broader research is undertaken to
continue learning about the reasons why noncompliance occurs.

o  Establish a long-term, quantitative voluntary compliance goal for

  Agency Comments
  and Our Evaluation

individual income tax underreporting and for tax underpayment, as well as
for other areas of noncompliance as data become available.

The Commissioner of Internal Revenue provided written comments on a draft
of this report in a letter dated July 6, 2005, which is reprinted in
appendix IV. In the letter, the Commissioner agreed with our
recommendations. In response to the recommendation that IRS develop plans
to periodically measure tax compliance, the Commissioner recognized the
need for and value of developing and regularly updating compliance
measures for various taxpayer populations and said that IRS will continue
to consult with stakeholders to develop and refine its compliance
measurement plans. In response to our recommendation that IRS take steps
to regularly collect complete, accurate, and consistent data on the
reasons for noncompliance, the Commissioner agreed that a better
understanding of taxpayer noncompliant behavior would be useful in shaping
strategic priorities and defining efforts to improve compliance. He
further said that the operating divisions will continue to partner with
the IRS research community to identify and better understand specific
reasons for noncompliance and that IRS will ensure that auditors are
trained to properly apply reason codes in the new report-writing system
IRS is developing. In response to the recommendation that IRS develop
longterm quantitative compliance goals, the Commissioner agreed with the
concept of developing such goals and discussed factors that make
goalsetting challenging. We appreciate IRS's current actions related to
our recommendations and recognize the challenges involved in balancing a
number of complex issues related to obtaining and using tax compliance
data.

As agreed with your offices, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 30 days from
its issue date. At that time, we will send copies to the Chairman and
Ranking Minority Member, House Committee on Ways and Means; the Secretary
of the Treasury; the Commissioner of Internal Revenue; the Director,
Office of Management and Budget; and other interested parties. We will
make

copies available to others on request. In addition, the report will be
available at no charge on the GAO Web site at http://www.gao.gov/.

If you or your staff have any questions, please contact me at (202)
512-9110 or [email protected]. Contact points for our Offices of
Congressional Relations and Public Affairs may be found on the last page
of this report. Key contributors to this report are listed in appendix V.

Michael Brostek Director, Tax Issues Strategic Issues

Appendix I: IRS Compliance Measurement Surveys

The following table summarizes the Internal Revenue Service's (IRS)
efforts to measure voluntary compliance using TCMP surveys and the
National Research Program (NRP) survey of individual income tax returns
for tax year 2001. Years provided for individual income tax surveys refer
to tax years. Years provided for surveys for all other types of tax refer
to return processing years.

               Table 3: Types of Surveys by Return Type and Year

                       Page 30 GAO-05-753 Tax Compliance
                                                                                                                                                                                                                                                                                  Exempt                                                           Employee                                               S              Delinquent                                                                                    
Return Year Sample Individual 1963 92,000  1965 50,000  1969 53,000  1971 26,000  1973 55,000  1976 50,000  1979 55,000  1982 50,000  1985 50,000  1988 54,000  2001 46,000        Small 1969 16,000  1973 20,000  1978 33,000  1981 33,000  1988 19,000  Estate 1971 4,600 organization 1974 11,400  1979 20,000  1988 3,000 Fiduciary 1975 8,900     plan 1982 18,000 Partnership 1982 27,000 corporation 1985 10,000 returns-non 1963 27,000  1966 114,000  1969 70,000         Delinquent 1979 25,000       
  type        size income tax                                                                                                                                               corporations                                                                 returns                 returns                                        returns             returns                 returns                 returns                    farm                                        returns-individual                   
                                                                                                                                                                                                                                                                                                                                                                                                                           business                                                                        1988 25,000

Appendix I: IRS Compliance Measurement Surveys

                                    Return type        Year       Sample size 
                 Surveys of delinquent accounts        1963           178,000 
                                                       1964           166,000 
                                                       1969         1,800,000 
                                                       1970         1,800,000 
                                                       1971         1,800,000 
                                                       1981         1,800,000 
                                                       1984         1,800,000 

Source: GAO, Tax Administration: IRS' Plans to Measure Tax Compliance Can
Be Improved, GAO/GGD-93-52 (Washington, D.C.: Apr. 5, 1993); IRS,
Understanding the Tax Gap, FS-2005-14, (March 2005).

Appendix II: Detailed Tax Gap Estimates, Data Sources, and Level of Certainty

The following table shows estimates for the various portions of the
preliminary 2001 tax gap, the sources, including the age, of the data the
Internal Revenue Service (IRS) used for these estimates, IRS's level of
certainty for each estimate, and areas for which IRS could not develop an
estimate because of insufficient data.

Table 4: IRS's Preliminary Tax Year 2001 Gross Tax Gap Estimates, Data
Sources, and Level of Certainty by Tax Gap Component and Type of Tax

                            Estimate                       
Tax gap component &       dollars                       
       type of tax     (in billions)     Estimate data          IRS certainty 
                                           source(s)                    level 
     Underreporting      $250-$292                                 -          

                                       c

Individual income tax $150-$187  o  Tax Year (TY) 2001 National Research
Program (NRP)

o  TY 1988 and earlier TCMP studies

o  	1981 and 1985-6 University of Michigan surveys of consumers (informal
suppliers)

o  	1984 University of Illinois study of restaurants and other eating
places (tip income)

                                       -

                                       c

                                       c

                           Business income $83-$99 -

                                       c

                         Non-business income $42-$57 -

                                       c

Adjustments, deductions, exemptions

                                   $14-$16 -

                                       c

Credits $11-$14 -

                       Corporation income tax $30 Weaker

Large corporations $25 o   Operational audits averaged over 1984,   Weaker 
                                           1985, & 1986                
                          o  TY 1982 TCMP study of unrelated business  
                                           income tax of               
                                     tax-exempt organizations          
                          o      TY 1975 TCMP study on fiduciaries     

Small corporations $5  o  TY 1977 and 1980 TCMP surveys Weaker

                                       c

                            Employment tax $66-$71 -

Self-Employment tax $51-$56  o  TY 2001 NRP

o  TY 1984 withholding noncompliance study

o  	1981 and 1985-6 University of Michigan surveys on informal suppliers

                                       -

                                       c

Employer-withheld $14  o  1984 University of Illinois study on tip income
Weaker
employment tax  o  TY 1984 withholding noncompliance study
(FICA)

 Unemployment tax $1  o  1984 University of Illinois study on tip income Weaker

o  TY 1984 withholding noncompliance study

Estate tax $4  o  	IRS's Statistics of Income (SOI) associated with filed
estate Reasonable tax returns for TY 1992

Excise tax no estimate N/A N/A

Appendix II: Detailed Tax Gap Estimates, Data Sources, and Level of
Certainty

                               Estimate                         
    Tax gap component &         dollars                         
        type of tax       (in billions) Estimate data source(s) IRS certainty 
                                                                        level 
       Underpaymenta              $31.7     IRS Master File        Actual     
Individual income tax          $19.4     IRS Master File        Actual     
Corporation income tax          $2.3     IRS Master File        Actual     
       Employment tax              $7.2     IRS Master File        Actual     
         Estate tax                $2.3     IRS Master File        Actual     
         Excise tax                $0.5     IRS Master File        Actual     
         Nonfilingb                 $30                            Reasonable 
Individual income tax            $28   o  TY 1988 Nonfiler      Reasonable 
                                                 TCMP           
Corporation income tax  no estimate            N/A                N/A      
       Employment tax      no estimate            N/A                N/A      

Estate tax $2  o  	2 University of Michigan longitudinal surveys (begun in
1992 Reasonable and 1993 and interviews participants every 2 years)

o  TY 1992 IRS's SOI

Excise tax no estimate N/A N/A

                                Total $312-$353

Source: IRS.

Notes: Figures may not sum to totals due to rounding. N/A = not available.

a

Unlike the other components of the 2001 tax gap, the underpayment
component is not an estimate, but rather represents the tax amounts that
taxpayers reported on time, but did not pay on time.

b IRS's nonfiler estimate for individual income tax is net of amounts of
true tax liability that are paid on time (e.g., through withholding).
However, refunds that are due to nonfilers do not reduce the nonfiling
gap, since they are not associated with a tax liability.

c

These estimates are based on more recent NRP data, but IRS has not
finalized the certainty level for these estimates because it has not yet
completed its assessment of the quality of the NRP data.

Appendix III: IRS Key Efforts to Reduce the Tax Gap

The Internal Revenue Service's (IRS) strategic plan outlines, but does not
prioritize, service and enforcement efforts to improve compliance.
Therefore, we asked IRS officials to identify IRS's key efforts to reduce
the tax gap. IRS's divisions provided lists that totaled 47 efforts, which
are described in the following examples.

The Small Business/Self-Employed Division identified 15 efforts such as
models to identify higher priority collection cases to pursue, a computer
matching program to identify underreported income, initiatives on high
income nonfilers, attempts to improve tip income reporting, and efforts to
identify abusive tax avoidance transactions.

The Wage and Investment Division identified 7 efforts including various
initiatives on tax collection, Earned Income Tax Credit, and using private
contractors to collect certain types of tax debts.

The Large and Mid-Sized Business Division identified 5 efforts such as
identifying compliance risks, starting examinations sooner and doing them
faster, and improving the treatment of abusive tax avoidance transactions.

The Tax Exempt and Government Entities Division identified 8 efforts
including abusive tax avoidance transactions in employee plans, abuses in
tax-exempt bond financing, pension plan noncompliance, and abuses by
credit counseling organizations.

The Criminal Investigation Division identified 12 efforts including those
involving questionable refunds, nonfilers, employment tax evasion,
corporation fraud, and offshore abusive tax schemes.

Appendix IV: Comments from the Internal Revenue Service

Appendix IV: Comments from the Internal Revenue Service

Appendix IV: Comments from the Internal Revenue Service

Appendix IV: Comments from the Internal Revenue Service

Appendix V: GAO Contact and Staff Acknowledgments

  GAO Contact Acknowledgments

(450412)

Michael Brostek, (202) 512-9110

In addition to the contact named above, Jeff Arkin, Ralph Block, Elizabeth
Curda, Elizabeth Fan, Evan Gilman, Shannon Groff, George Guttman, Michael
Rose, Sam Scrutchins, and Tom Short made key contributions to

this report.

  GAO's Mission

Obtaining Copies of GAO Reports and Testimony

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

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

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

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

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

  To Report Fraud, Contact:

Waste, and Abuse in Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: [email protected] Programs Automated answering system: (800)
424-5454 or (202) 512-7470

Gloria Jarmon, Managing Director, [email protected] (202)
512-4400Congressional U.S. Government Accountability Office, 441 G Street
NW, Room 7125 Relations Washington, D.C. 20548

Public Affairs 	Paul Anderson, Managing Director, [email protected] (202)
512-4800 U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548

                           PRINTED ON RECYCLED PAPER
*** End of document. ***