Tax Administration: IRS Measures Could Provide a More Balanced Picture of
Audit Results and Costs (Letter Report, 06/23/98, GAO/GGD-98-128).

Pursuant to a congressional request, GAO provided information on the
Internal Revenue Service's (IRS) measures of the results of its audits
of tax returns, focusing on: (1) how much of the additional taxes
recommended in all types of audits that were closed in fiscal year (FY)
1992 through FY 1997 had been settled or were still in dispute and, if
settled, how much had been assessed and collected as of September 27,
1997; (2) how much of the recommended additional taxes had been assessed
and collected for audits closed in FY 1992; and (3) whether broad IRS
measures of audit results fully represented audit revenues and costs.

GAO noted that: (1) for audits closed in FY 1992 through FY 1997, IRS
recommended tens of billions of dollars in additional taxes for each
year; (2) however, not all recommended taxes are assessed; and not all
assessed taxes are collected; (3) IRS had settled 40 percent of FY 1992
audits without assessing the recommended taxes, usually because of IRS
Office of Appeals' decisions, and had yet to settle the assessment
status of the other 26 percent; (4) of the $8.5 billion assessed, IRS
had collected 72 percent, which means that 25 percent of all recommended
taxes for FY 1992 audits had been collected as of September 27, 1997;
(5) for audits closed in FY 1993 through FY 1997, assessment and
collection results were less complete because less time had elapsed for
these actions to occur compared with 1992; (6) the assessment and
collection rates varied by the type of audit closed in FY 1992; (7) in
general, IRS assessed a higher percentage of the assessed taxes for
simpler audits compared with complex audits; (8) however, IRS collected
a higher percentage of the recommended taxes from the simpler audits
than complex audits; (9) for simpler service center audits, IRS had
assessed 76 percent of the recommended taxes but had collected 56
percent of the assessed taxes as of September 27, 1997; (10) at the
other extreme, after audits of complex returns from Coordinated
Examination Program (CEP) corporations, IRS had assessed 20 percent of
the recommended taxes but had collected 97 percent of the assessed
taxes; (11) as of September 27, 1997, 39 percent of the amounts
recommended in CEP audits were still in dispute; (12) IRS' existing
performance measures do not cover all audit-based revenues or costs;
(13) measuring the taxes recommended does not account for the related
assessments and collections; nor does it account for indirect revenue
gains; (14) measuring other types of revenues are important because not
all recommended taxes are assessed or collected; (15) IRS measures the
staff time directly charged to audits but not the dollar costs of this
direct time; (16) compiling complete and reliable data on the indirect
revenues and taxpayer costs can be very difficult to do because of
limitations in the data sources; (17) beyond these limitations, IRS did
not use other existing data to develop and report measures that more
fully represented audit results; (18) for additional measure, audit
revenues could be compared with related costs; and (19) to develop such
measures, IRS would need more data on both direct and indirect costs.

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

 REPORTNUM:  GGD-98-128
     TITLE:  Tax Administration: IRS Measures Could Provide a More 
             Balanced Picture of Audit Results and Costs
      DATE:  06/23/98
   SUBJECT:  Tax administration
             Tax return audits
             Government collections
             Income taxes
             Data integrity
             Federal taxes
             Cost analysis
IDENTIFIER:  IRS Coordinated Examination Program
             IRS Enforcement Revenue Information System
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


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


Report to the Honorable
Rob Portman, Committee on Ways and Means, House of Representatives

June 1998

TAX ADMINISTRATION - IRS MEASURES
COULD PROVIDE A MORE BALANCED
PICTURE OF AUDIT RESULTS AND COSTS

GAO/GGD-98-128

IRS' Audit Results and Costs Measures

(268807)


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

  AIMS - Audit Information Management System
  CEP - Coordinated Examination Program
  ERIS - Enforcement Revenue Information System
  IRS - Internal Revenue Service

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


B-277433

June 23, 1998

The Honorable Rob Portman
Committee on Ways and Means
House of Representatives

Dear Mr.  Portman: 

The Internal Revenue Service's (IRS) mission is to collect the proper
amount of taxes at the least cost to the federal government and
taxpayer.  As part of its effort to achieve its mission, IRS audits
all types of tax returns to determine whether taxpayers have reported
their correct tax liabilities.  IRS spends over a billion dollars
annually on its audit programs, including staff costs.  These audits
also impose costs and burdens on taxpayers.  On the basis of its
audits, IRS annually recommends that taxpayers pay additional taxes
totaling tens of billions of dollars. 

IRS has traditionally measured the overall results of audits by the
amount of recommended additional taxes and by the amount of direct
staff time used in an audit.  In 1994, we reported that these
measures for large corporations provided an incomplete picture of
audit results, because much of the recommended taxes had not been
assessed and collected after being disputed by taxpayers.\1 As a
result, we recommended that IRS expand its measures for audits of
large corporations to include the percent of recommended taxes that
is ultimately collected.  In its budget request to Congress for
fiscal year 1999, IRS continues to report audit revenues only as
dollars recommended. 

This report responds to your request for information about IRS'
measures of the results of its audits of tax returns.  Our objective
was to determine how much of the additional taxes recommended in all
types of audits that were closed in fiscal years 1992 through 1997
had been settled or were still in dispute and, if settled, how much
had been assessed and collected as of September 27, 1997.  In
addition, for audits closed in fiscal year 1992, our objectives were
to determine how much of the recommended additional taxes had been
assessed and collected across all types of audits and whether broad
IRS measures of audit results fully represented audit revenues and
costs. 


--------------------
\1 Tax Administration:  Compliance Measures and Audits of Large
Corporations Need Improvement (GAO/GGD-94-70, Sept.  1, 1994). 


   BACKGROUND
------------------------------------------------------------ Letter :1

IRS has a multistage process that governs audits and settlements of
disputes over the additional taxes recommended.  In an audit, an
auditor, usually from IRS' Examination Division, is to review a
taxpayer's books and records to determine compliance with tax laws in
reporting the proper amount of tax.  Auditors usually recommend
additional tax assessments but may recommend a decrease or no change
in the tax reported on the return, depending on the documentation
provided by the taxpayer.  If the taxpayer agrees to pay or does not
respond to IRS' notices on recommended additional taxes, IRS assesses
the tax--that is, formally notifies the taxpayer that the specified
amount of tax is owed and that interest and penalties may accrue if
the tax is not paid by a certain date. 

Taxpayers who do not agree with the recommended additional taxes can
(1) file a protest with the IRS Office of Appeals, (2) take the
dispute to tax court without paying the recommended tax, or (3) pay
the tax and claim a refund in the U.S.  Court of Federal Claims or a
federal district court.  Of these options, taxpayers usually protest
to IRS Appeals.\2 Appeals settles most of these disputes, and the
remainder are docketed for trial.  If Appeals is unsuccessful in
settling the dispute, the Office of Chief Counsel gets involved in
settlement as well as in any trial.  The agreements made in
settlements and the rulings made in trials dictate how much of the
disputed amount gets assessed.  The assessed amount--not the
recommended amount--establishes the taxpayer's liability.  If
taxpayers do not pay the taxes that are assessed, IRS can take action
to collect the taxes. 

IRS tracks the additional taxes recommended by audit classes, which
are based on the amount of reported income or assets and type of
return.  Across the audit classes, tax returns vary in complexity,
ranging from simple individual returns to complex corporation
returns.  The classes include specialized audit programs, such as the
Coordinated Examination Program (CEP) for the nation's largest
corporations. 

The range of size and complexity across tax returns affects the
amount of time and resources IRS uses to audit a return and resolve
disputes over the assessment of recommended taxes.  Audits of large
corporations usually take 2 to 3 years.  If the large corporation
disputes the recommended additional taxes, another 2 to 3 years can
elapse in trying to settle the dispute through Appeals; several
additional years may be needed if the dispute goes to trial.  For
smaller, less complex returns, the time IRS uses to audit and settle
any disputes over recommended tax assessments is shorter; but these
processes usually take at least 1 year. 

To deal with this varying complexity, IRS has three types of
auditors.  First, lower graded tax examiners at IRS service centers
audit simpler individual returns through correspondence.  Second,
higher graded revenue agents from IRS district offices visit
individuals, corporations, and other types of taxpayers to address
the most complex tax returns; they also work in teams to audit CEP
and some other large corporations.  Third, tax auditors, who usually
do audits by meeting with taxpayers at a district office, fall in the
middle ranges regarding their grades and the complexity of their
audits.  IRS' appeals officers also differ by their grades and by the
complexity and size of their workloads. 


--------------------
\2 Taxpayers also may appeal audits that recommend decreases or no
changes to the reported tax when they believe that their tax
liabilities should be even less. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :2

For audits closed in fiscal years 1992 through 1997, IRS recommended
tens of billions of dollars in additional taxes for each year.  IRS
reported these recommended taxes to Congress as its audit results. 
However, not all recommended taxes are assessed; and not all assessed
taxes are collected.  For example, as a result of the fiscal year
1992 audits, IRS had assessed 34 percent ($8.5 billion) of the $24.8
billion in recommended additional taxes as of September 27, 1997. 
IRS had settled another 40 percent ($9.9 billion) without assessing
the recommended taxes, usually because of Appeals' decisions, and had
yet to settle the assessment status of the other 26 percent ($6.4
billion).  Of the $8.5 billion assessed, IRS had collected 72 percent
($6.1 billion), which means that 25 percent of all recommended taxes
for fiscal year 1992 audits had been collected as of September 27,
1997.  For audits closed in fiscal years 1993 through 1997,
assessment and collection results were less complete because less
time had elapsed for these actions to occur compared with 1992. 

The assessment and collection rates varied by the type of audit
closed in fiscal year 1992.  In general, IRS assessed a higher
percentage of the recommended taxes but collected a lower percentage
of the assessed taxes for simpler audits compared with complex
audits.  However, IRS collected a higher percentage of the
recommended taxes from the simpler audits than complex audits.  For
simpler service center audits, IRS had assessed 76 percent of the
recommended taxes but had collected 56 percent of the assessed taxes
as of September 27, 1997.  As of that date, all of the service center
audits had been settled.  At the other extreme, after audits of
complex returns from CEP corporations, IRS had assessed 20 percent of
the recommended taxes but had collected 97 percent of the assessed
taxes.  As of September 27, 1997, 39 percent of the amounts
recommended in CEP audits were still in dispute.  Overall, IRS had
collected less than half of the recommended taxes across all types of
audits--ranging from 20 percent (CEP) to 43 percent (service center). 

IRS' existing performance measures do not cover all audit-based
revenues or costs.  Measuring the taxes recommended does not account
for the related assessments and collections; nor does it account for
indirect revenue gains, such as when audits enhance voluntary
compliance.  Measuring other types of revenues is important because
not all recommended taxes are assessed or collected.  IRS measures
the staff time directly charged to audits but not the dollar costs of
this direct time.  Likewise, IRS does not measure indirect costs
incurred by the Examination Division for training, office space, and
the like or the direct and indirect costs that audits create outside
of the Examination Division for IRS as well as for taxpayers. 
Compiling complete and reliable data on the indirect revenues and
taxpayer costs can be very difficult to do because of limitations in
the data sources and research. 

Beyond these limitations, IRS did not use other existing data to
develop and report measures that more fully represented audit
results.  For example, in reporting audit results, data on taxes
recommended could be balanced with data on taxes assessed and
collected.  If these existing assessment and collection data were
reported, fuller insights would be available on IRS' audit revenues. 
For additional measures, audit revenues could be compared with
related costs.  To develop such measures, IRS would need more data on
both direct and indirect costs.  Given the problems inherent in
collecting and analyzing data on indirect costs, the role of
management judgment in making decisions about audits and audit
resources will continue to be essential. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :3

To address our objectives on the assessments, collections, and costs
related to taxes recommended in audits, we used IRS' Enforcement
Revenue Information System (ERIS) data as of September 27, 1997.\3
Our analyses started with audits closed in fiscal year 1992. 
According to IRS officials, ERIS data were unreliable prior to 1992. 
We focused most of our analyses on 1992, instead of later years, to
allow the most time for assessment and collection actions to have
been taken on additional amounts recommended in audits.  Because ERIS
attempts to continually update the status of these recommended
amounts, ERIS' data could differ from data in other IRS systems that
are not doing such updating. 

We did not test the reliability of the data provided to ERIS by other
IRS systems or data processed by ERIS.\4 However, we followed up with
IRS on any anomalies that we found in the data and adjusted the data
or our methodology as needed.  For example, as a result of
questioning by us and other IRS officials, ERIS officials discovered
a mistake during April 1998 in the documentation provided to us for
interpreting ERIS data on the settlement status of recommended taxes. 
Subsequent discussions with ERIS officials clarified the approach to
use for interpreting the ERIS data.  IRS officials concurred with the
analytical results generated from using this approach. 

To determine how much of the recommended additional tax amounts had
been settled, assessed, and collected as of September 27, 1997, for
audits closed in fiscal years 1992 through 1997, we matched the
recommended amounts to actions taken in Appeals or the Office of
Chief Counsel.\5 In calculating the final assessments, we subtracted
amounts that IRS initially assessed from the additional taxes
recommended but later abated.\6

Then, for each fiscal year, we added the assessments and collections
that took place in each IRS function--Examination, Appeals, and Chief
Counsel--to obtain the overall results for the audits.  We arrayed
these results by the fiscal year of the audit closure to develop a
broad picture of what happened to recommended tax amounts.  (App.  II
shows the results for each fiscal year for seven types of audits.)

To determine how much of the recommended additional tax amounts had
been assessed and collected as of September 27, 1997, by type of
audit closed in fiscal year 1992, we did analyses similar to those
done for our first objective.  We expanded the analyses to array the
results by seven types of audits.  We developed these 7 types from 30
subcategories that IRS used to classify audits by type of tax return,
tax, or taxpayer (app.  III shows the subcategories).\7 The seven
types include income tax audits at IRS service centers or at IRS
district offices of nonbusiness individuals, business individuals,
small corporations, non-CEP large corporations, and CEP corporations
as well as those audited for other types of tax liabilities.\8 At the
request of IRS officials, we separately analyzed service center and
CEP audits because of their specialized natures. 

To determine whether IRS' broad measures of audit results fully
represented audit revenues and costs, we reviewed IRS' budget
submissions for fiscal years 1998 and 1999 to see the data reported
as audit results and performance measures.  We analyzed available
ERIS data on the revenues and costs associated with audits to
identify audit measures that could be developed.  We then created
three ratios of tax revenues to costs for audits closed in fiscal
year 1992.  These ratios compared additional taxes recommended,
assessed, and collected with the related direct costs through each of
these stages.  We arrayed these three ratios across the seven types
of audits. 

For the revenues, we used available IRS data on the additional taxes
recommended, assessed, and collected.  For the costs, we used
available data on direct time charged by staff who do the audits and
settle disputes over the additional taxes recommended;\9 data on the
direct staff time to collect the additional taxes assessed were not
available.  We identified the direct hours charged by staff grade
level in Examination, Appeals, and Chief Counsel.\10 We then applied
an hourly rate to the grade levels by using the General Schedule pay
tables for 1992 through 1997.  For each grade, we adjusted the
midpoint of the pay scale to account for locality pay.  We accounted
for work hours available and costs of staff benefits, such as paid
vacation and sick leave, by using Office of Management and Budget
Circular No.  A-76.\11 Using these factors, we computed the direct
staff costs. 

We discussed our methodology and our results with IRS officials who
manage ERIS or who manage the audit, settlement, and collection
activities.  We incorporated their suggestions into our work as
appropriate.  We requested comments on a draft of this report from
the Commissioner of Internal Revenue.  His comments are discussed
near the end of this letter and reprinted in appendix V.  We
performed our audit work at IRS' National Office in Washington, D.C.,
between July 1997 and May 1998 in accordance with generally accepted
government auditing standards. 


--------------------
\3 Since 1990, IRS has used ERIS to track actions taken outside of
the Examination Division on recommended additional taxes (app.  I
provides further information on ERIS). 

\4 These systems include the individual and business masterfiles, the
Audit Information Management System, and other systems on IRS
enforcement actions. 

\5 In addition to recommending additional taxes, IRS officials said
audits produce revenue in other ways.  For example, IRS audits
"protect" revenue (e.g., disallow claims filed by taxpayers for
refunds of previously assessed taxes) and generate interest on
additional assessments.  Our analyses excluded protected revenue
because our objective was to track the recommended additional taxes. 
We excluded interest because it is a function of time and is not
recommended by Examination. 

\6 Auditors can recommend (1) additional taxes, (2) reductions in
taxes, or (3) no changes in taxes.  The audit results in this report
do not include recommended reductions in taxes because our objectives
dealt with actions taken on the recommended additional taxes. 

\7 The categories used by IRS to classify audits may change if IRS is
restructured as proposed by the Commissioner of IRS. 

\8 The difference between nonbusiness and business individuals
depends on whether most of the reported income derives from
self-employment in a business.  The difference between a small and
large non-CEP corporation depends on whether a corporation reports
less than $10 million in assets.  The large corporation category also
includes audits of foreign corporations.  The "other" category
includes all other tax liabilities, such as employment, gift, and
estate taxes. 

\9 Direct time charged by staff excludes time charges for such things
as training or administration. 

\10 In computing the direct costs, we included audits that resulted
in no change or in a reduction to the tax reported on the return.  We
reasoned that IRS incurred these costs in an attempt to identify
taxpayers owing additional tax amounts. 

\11 When we began our review, ERIS data did not include the costs of
staff benefits.  We shared our computation of these costs with IRS
officials.  ERIS now includes the costs of staff benefits. 


   IRS ASSESSED AND COLLECTED ONLY
   PART OF THE TAXES RECOMMENDED
------------------------------------------------------------ Letter :4

IRS annually reports to Congress on the amount of additional tax and
penalties recommended in audits closed in each fiscal year.  For the
years we reviewed, the recommended amounts ranged up to about $32
billion.  However, the recommended amount does not represent the
actual revenue resulting from IRS' audits.  As of September 27, 1997,
less than half of the recommended amount had been assessed as
additional taxes, and not all of the assessed amount had subsequently
been collected. 

For fiscal year 1992 audits, as of September 27, 1997, IRS had
assessed $8.5 billion (34 percent) of the $24.8 billion in
recommended additional taxes.  IRS settled another 40 percent of the
recommended additional taxes without ultimately assessing the
recommended additional taxes for various reasons.  These reasons
included taxpayer claims; decisions in Appeals or the courts; and
reductions (i.e., abatements) of amounts that had initially been
assessed.\12 Disputes on the final 26 percent of the additional
recommended taxes had yet to be settled.  As of September 27, 1997,
IRS had collected $6.1 billion--or 72 percent of the additional taxes
assessed and 25 percent of the additional taxes recommended.  Table 1
shows these results for audits closed in fiscal year 1992 as well as
for fiscal years 1993 through 1997. 



                                      Table 1
                      
                      Status of Additional Amounts Recommended
                       for Audits Closed in Fiscal Years 1992
                       Through 1997, as of September 27, 1997

                              (Dollars in millions\a)

                Settlement status of recommended
                            amounts                      Collection status
                --------------------------------  --------------------------------
Fi
sc                                                                      As percent
al  Recommende  Percent of     Percent            Collecte  As percent   of amount
ye           d    disputes         not   Percent         d   of amount  recommende
ar   amounts\b   unsettled  assessed\b  assessed   amounts    assessed           d
--  ----------  ----------  ----------  --------  --------  ----------  ----------
19     $24,833         26%         40%       34%    $6,138         72%         25%
 92
19      22,022          25          35        40     6,022          68          27
 93
19      22,637          31          28        41     6,059          65          27
 94
19      27,184          41          19        40     6,878          63          25
 95
19      30,771          46          18        36     6,737          61          22
 96
19      31,691          55           7        38     6,447          54          20
 97
----------------------------------------------------------------------------------
Note:  Appendix II shows these results for each fiscal year for the
seven types of audits. 

\a Dollars are in current dollars. 

\b Recommended amounts are considered "not assessed" when (1) they
are not sustained during IRS' appeals process; (2) they are sustained
during appeals but offset by taxpayers' claims, such as net operating
losses from other tax years, before being assessed; and/or (3) they
were initially assessed but later abated because of information or
claims filed by the taxpayers. 

Source:  GAO analysis of IRS ERIS data. 

Since 1992, the results of the audits were less complete because
larger percentages are unsettled for each succeeding year, reaching
55 percent in 1997.  As more settlements occur over time, the rates
at which the recommended taxes are assessed and collected should
increase.  IRS officials said they believe that this assessment rate
is higher in more recent years compared with 1992 because IRS has
been trying to obtain taxpayer agreement with any taxes recommended
before the audit is closed.  However, it is not yet clear whether the
rate at which assessed amounts are collected for 1993 through 1997
audits will exceed the rate for 1992.  Because of the incomplete
results for audits closed in these more recent years, our analyses
focused on 1992 audits. 


--------------------
\12 Certain factors can reduce or eliminate assessments from
audit-based recommendations.  For example, (1) prior to assessment,
recommended taxes that were upheld by Appeals can be offset by
taxpayers' claims, such as those for net operating losses carried
over from other tax years and (2) after the audit and appeals
processes, taxpayers may provide additional information that can
offset the taxes assessed.  IRS' databases cannot easily track all
these factors.  In our report GAO/GGD-94-70, we recommended that IRS
improve its databases to begin accounting for such factors. 


   COLLECTION AND ASSESSMENT RATES
   VARY BY TYPE OF AUDIT
------------------------------------------------------------ Letter :5

Table 2 shows how much of the recommended additional taxes had been
settled, assessed, and collected as of September 27, 1997, for seven
types of audits closed in fiscal year 1992.  In general, the
assessment and collection rates varied by the complexity of the
audit.  With more complex audits, such as corporation audits,
taxpayers were more likely to reduce additional assessments by
appealing amounts recommended but also more likely to pay almost all
of the amounts assessed.  For example: 

  -- Assessment rates for the simpler audits done at service centers
     and for the audits of individuals were higher than for corporate
     audit categories.  The rates were 76 percent for service center
     audits and about 59 percent for audits of individuals, but the
     rates were 20 percent for CEP audits and 33 percent for other
     large corporation audits. 

  -- The rates for collection of assessed amounts were higher for
     corporation audits.  IRS collected about half of the assessed
     amounts for service center and individual audits compared with
     97 percent for CEP and 73 percent for other large corporation
     audits. 

The percentage of recommended amounts that IRS collects is also
affected by these assessment and collection rates.  For fiscal year
1992 audits, the percentage of recommended taxes that IRS had
collected as of September 27, 1997, ranged from 20 percent for CEP
audits to 43 percent for service center audits. 



                                     Table 2
                     
                     Status of Additional Amounts Recommended
                     for Fiscal Year 1992 Across Seven Types
                       of Audits, as of September 27, 1997

                              (Dollars in millions)

                     Settlement status of
                    recommended amounts\a              Collection status
                ------------------------------  --------------------------------
Ty
pe
of                           Percent                                  As percent
Au              Percent of       not                      As percent   of amount
di  Recommende    disputes  assessed   Percent  Collecte   of amount  recommende
t    d amounts   unsettled        \b  assessed  d amount    assessed           d
--  ----------  ----------  --------  --------  --------  ----------  ----------
Se      $1,464          0%       24%       76%     $ 625          56         43%
 r
 v
 i
 c
 e
 c
 e
 n
 t
 e
 r
In       2,763           4        37        59       894          55          32
 d
 i
 v
 i
 d
 u
 a
 l
 n
 o
 n
 b
 u
 s
 i
 n
 e
 s
 s
In       1,290           6        33        60       371          48          29
 d
 i
 v
 i
 d
 u
 a
 l
 b
 u
 s
 i
 n
 e
 s
 s
Sm       1,001           6        42        52       301          58          30
 a
 l
 l
 c
 o
 r
 p
 o
 r
 a
 t
 i
 o
 n
 s
La       2,221           7        60        33       543          73          24
 r
 g
 e
 c
 o
 r
 p
 o
 r
 a
 t
 i
 o
 n
 s
CE     13, 893          39        41        20     2,757          97          20
 P
Ot       2,202          27        34        40       646          74          29
 h
 e
 r
 \
 c
================================================================================
To      24,833          26        40        34     6,138          72          25
 t
 a
 l
 \
 d
--------------------------------------------------------------------------------
Note:  Table III.1 shows this information by the 30 selected
categories

\a Percentages do not total to 100 percent due to rounding. 

\b Recommended amounts are considered "not assessed" when (1) they
are not sustained during IRS' appeals process; (2) they are sustained
during appeals but offset by taxpayers' claims, such as net operating
losses from other tax years, before being assessed; and/or (3) they
were initially assessed but later abated because of information or
claims filed by the taxpayers. 

\c "Other" includes audits of returns for employment tax, estate tax,
excise tax, and gift tax; audits conducted in IRS training; and
audits categorized by IRS as other. 

\d Amounts do not add to totals due to rounding. 

Source:  GAO analysis of IRS ERIS data. 

We asked IRS officials to explain differences in the assessment and
collection rates across the types of audits.  IRS officials said
several factors contributed to these variances.  For example,
comparing the service center and corporation audits, they said: 

  -- The rate for assessing the recommended amount is higher for
     service center audits, because the taxpayers are (1) more likely
     to not respond to IRS correspondence and notices, which leads
     IRS to assess the recommended amounts and (2) less likely to
     dispute the recommended amounts, usually much smaller and
     involving simpler tax issues, compared with audits of
     corporations for which disputes are much more likely. 

  -- The rate for collecting the assessed amount is lower for service
     center audits because the taxpayers tend to have fewer assets to
     pay the tax assessment.  Further, CEP and large corporations are
     more likely to pay the assessments to avoid large interest
     charges.  Small corporations and individuals with businesses
     also may have difficulties finding the funds to pay the
     additional tax assessments. 


   IRS' PERFORMANCE MEASURES DO
   NOT FULLY REFLECT ALL AUDIT
   REVENUES AND COSTS
------------------------------------------------------------ Letter :6

IRS' performance measures do not fully reflect all audit-based
revenues and costs.  For example, IRS' existing measures on taxes
recommended do not include indirect revenues resulting from the
effect of audits on voluntary compliance.  Although IRS measures the
staff time directly spent on audits, it does not measure the dollar
costs of this direct time or the indirect costs incurred by the
Examination Division for such things as training time or office
space.  IRS also does not measure direct and indirect costs that
audits create outside of the Examination Division for IRS as well as
taxpayers.  Compiling complete and reliable data on the indirect
revenues and costs from outside IRS can be very difficult because of
limitations with the data sources and research. 

Further, IRS does not use its available data to develop and report
measures that would provide a fuller, more balanced picture of audit
results.  For example, data on taxes recommended could be balanced
with data on taxes assessed and collected in reporting audit results
and, as we previously recommended for audits of large corporations,
in developing additional performance measures.  In developing these
measures, such revenue data could be related to information on the
costs of audits.  In addition, IRS has the capacity to track more
data beyond the direct staff costs. 


      IRS DATA ON REVENUES FROM
      AUDITS ARE INCOMPLETE
---------------------------------------------------------- Letter :6.1

IRS data on taxes recommended, assessed, and collected do not
represent all revenues from audits.  Data are not available on the
indirect revenue effects from audits.  For example, when audits
induce both audited and unaudited taxpayers to be more voluntarily
compliant, tax collections increase indirectly.  Other indirect
revenue effects occur if audits adjust how much of a tax deduction
can be claimed in one tax year versus other years or reduce that
claim for future years.\13

IRS has difficulty measuring the indirect revenue effects because of
limitations with the data and research methods.  Such measurements
require (1) data that reflect the impact of audits and other IRS
activities on the compliance of individual taxpayers and (2) a
research methodology that reliably distinguishes the effect of audits
on voluntary compliance from other influencing factors.  IRS has
researched the indirect effects of audits but has yet to develop
reliable estimates because of these limitations. 

IRS data also do not include the marginal revenue effects from doing
audits.  Such marginal effects are the changes in total revenues that
result when IRS incrementally changes the number of audits in an
audit class.  As with indirect effects, developing data on marginal
effects can be challenging.  Without data on the indirect and
marginal effects, IRS and Congress cannot know the full impacts of
audits. 


--------------------
\13 If an audit reduces a depreciation claim by $100 but allows that
$100 to be claimed in future years, tax revenue would be higher for
the audited year but lower whenever the claims are allowed.  If an
audit reduces but does not eliminate a net operating loss claim, the
audit would produce no tax revenue for the audited year but higher
tax revenue in the future when, because of the audit, a smaller loss
is claimed. 


      IRS DATA ON COSTS OF AUDITS
      ARE INCOMPLETE
---------------------------------------------------------- Letter :6.2

IRS data on both its direct and indirect costs from doing audits are
incomplete.  IRS associates its direct costs with the time charged by
the staff who do the audits, settle the audit disputes, and collect
the audit assessments for specific types of audits.  IRS has data
available for computing the direct staff costs for the audit and
settlement activities.  However, IRS did not have such data for the
collection activity because IRS' Collection Division did not track
the time that staff spent trying to collect the additional
assessments arising from specific types of audits.  Our analyses of
the significance of the direct costs of collection were inconclusive
because of missing data.\14

Also, IRS did not have data on its indirect costs for an audit.  IRS
considers its indirect costs to include management time, training
time, space, and other support given to those who do audits, settle
audit disputes, and collect audit assessments.  Likewise, IRS does
not account for indirect costs outside IRS, such as those imposed on
the audited taxpayer or on society when a taxpayer evades tax
liabilities.\15 However, as with indirect revenue, collecting the
data for quantifying these external indirect costs is a difficult
task.  IRS has been working on ways to measure taxpayers' costs and
is starting to survey taxpayers on their satisfaction with the audit
process.  This survey does not gather data on taxpayers' costs but
may prove useful to IRS in deciding how to quantify these costs. 


--------------------
\14 Our analysis focused on the amount of time IRS took to collect
additional tax assessments arising from audits.  Appendix IV
describes our analysis and results. 

\15 If audits reduce tax evasion, the social costs involving the
efficiency and equity of the tax system also could be reduced.  Some
taxpayers choose investments or occupations that provide
opportunities to evade taxes.  With fewer evasion opportunities,
taxpayers may use their resources more efficiently elsewhere in the
economy.  Reducing evasion also could make the tax system more
equitable.  Tax evasion creates inequities when taxpayers with the
same tax liability pay different tax amounts.  Greater equity may
balance, at least in part, any increased costs imposed on audited
taxpayers. 


      IRS DOES NOT REPORT ALL
      EXISTING DATA ON AUDIT
      RESULTS
---------------------------------------------------------- Letter :6.3

Although IRS lacks data on all of the revenues and costs associated
with audits, it does have data that could be used to measure selected
revenues and costs.  However, IRS does not measure and report all its
existing data on audit revenues, such as additional taxes assessed
and collected on the basis of taxes recommended in audits.  Further,
IRS has not developed data and measures on the costs related to each
type of revenue. 

This report discusses broad, IRS-wide dollar measures of audit
results to track actions on any additional taxes recommended in
audits.  Such measures do not account for all aspects of audit
performance, such as the proper treatment of taxpayers and the
decision to recommend no change or a reduction to the tax liability
reported on a return.  These measures are not intended to be used to
evaluate the performance of individual IRS employees. 

Over the years, IRS has measured the overall results of audits done
in the Examination Division by the amount of additional taxes
recommended and time charged directly to an audit by Examination
staff.  These measures do not employ existing data that could more
fully represent the revenues and costs associated with audits. 
First, the measures do not report how much of the recommended taxes
are actually assessed and collected.  Second, the measures do not
report the dollar costs of the direct time charged to an audit or the
indirect costs for the audit, settlement, and collection activities. 

Although useful in some ways, measuring audit performance by just
taxes recommended and direct audit staff time presents an unbalanced
picture of the audit results.  Tables 1 and 2 show large differences
between the amounts recommended, assessed, and collected.  Taxpayers
dispute most of the additional recommended amounts, and settlement of
these disputes results in smaller amounts being assessed and
collected. 

Relying too heavily on additional taxes recommended as a measure of
audit results may create undesirable incentives.  Our previous work
on audits of large corporations has raised concerns that relying on
recommended taxes as a performance indicator may encourage auditors
to recommend taxes that would be unlikely to withstand taxpayer
challenge and thus not be assessed and collected.  To the extent that
this happens, audited taxpayers could be unnecessarily burdened.  We
recommended that IRS balance its measures of audit performance by
adding such measures as taxes ultimately collected. 

Further, the direct time charged to audits does not measure the
dollar costs.  In computing the costs of the direct time charges, one
must recognize that the pay grade levels of staff assigned to audits
vary by type of audit.  Further, the direct time charged in the
Examination Division excludes direct staff time charged in Appeals,
Chief Counsel, and Collection as well as indirect costs for the
audit, settlement, and collection activities. 

To illustrate the importance of developing a more complete set of
measures, we compared three ratios that each measured a type of audit
revenue with the related direct cost for seven types of audits.  We
calculated direct costs using hours charged by staff in Examination,
Appeals, and Chief Counsel.  The audit revenues included the amounts
recommended, assessed, and collected. 

Table 3 shows that these ratios differ widely.  Many factors affect
audit-based revenues and costs.  For example, recommended amounts are
affected by the number of audits and amount recommended per audit. 
The number of auditors as well as their time charges and pay grades
affect direct audit costs.  Service center auditors have the lowest
grades and charge the least amount of time per audit, while CEP
auditors have the highest grades and charge the most amount of time. 
Also, the costs of settlement in corporate audits are likely to be
higher than in other audits because corporations are more likely to
dispute recommended assessments.  As a result, reliance on a single
measure gives a less complete picture of audit results than relying
on all three measures. 



                                Table 3
                
                    Taxes Recommended, Assessed, and
                 Collected Compared With Related Direct
                Staff Costs for Audits Closed in Fiscal
                     1992, as of September 27, 1997

                             Recommended
                                 to cost    Assessed to   Collected to
Type of audit                    ratio\a   cost ratio\b   cost ratio\c
-------------------------  -------------  -------------  -------------
Service center                     197:1          104:1           59:1
Individual                          29:1           14:1            8:1
 nonbusiness
Individual business                 21:1           11:1            5:1
Small corporations                  15:1            7:1            4:1
Large corporations                  44:1           13:1           10:1
CEP corporations                   124:1           24:1           23:1
Other\d                             23:1            8:1            6:1
======================================================================
Total                               50:1           16:1           11:1
----------------------------------------------------------------------
Note:  Table III.2 shows this information by the 30 selected
categories. 

\a Includes Examination's direct staff costs but not related indirect
costs. 

\b Includes Examination's, Appeals', and Counsel's direct staff costs
but not related indirect costs. 

\c Includes Examination's, Appeals', and Counsel's direct staff costs
but not Collection's direct staff costs and related indirect costs. 

\d "Other" includes audits of returns for employment, estate, excise,
and gift taxes; audits done in IRS training; and audits categorized
by IRS as other. 

Source:  GAO analysis of IRS ERIS data. 

The ratios in table 3 should not be used as official measures of
results because they do not account for all costs.  If costs such as
Collection's direct staff costs and IRS' indirect costs could be
included, the ratios would be smaller, and the differences by type of
audit could change significantly.  Direct staff time accounts for
about half of all time charged by auditors; much of the remaining
time produces indirect costs.  The allocation of time could vary by
type of audit.  For example, service center audits may have a higher
proportion of indirect costs given IRS' reliance on automation and
nonaudit staff to help the auditors.  Further, because individuals
tend to pay their additional tax assessments more slowly than
corporations (see app.  IV), IRS would be likely to incur more costs
to collect additional assessments from individuals. 

IRS has plans to develop a measure that approximates the least
complete ratio of the three ratios.  In its 1999 budget submission,
IRS reported plans to develop a ratio of the taxes recommended to the
audit costs.\16 IRS officials said the audit cost side of the ratio
will come from an activity-based costing model that IRS is
developing. 

We talked to IRS officials about improving these ratios by capturing
more data on the related costs.  They said that tracking the direct
staff time to collect audit assessments cannot now be done but would
be possible if Collection Division staff began to report the time
spent on audit-based assessments.  These officials said IRS plans to
use formulas to allocate indirect costs, such as for administration
and rent, of the audit, settlement, and collection activities to the
types of audits.  They said that the model should, at a minimum, give
them a better basis for knowing more about the nature and magnitude
of the indirect costs. 

Further, we asked IRS about opportunities to more fully report
existing data on the revenues generated by audits.  Our interviews
with IRS officials indicated that measuring and reporting information
on taxes assessed and collected from audits, in addition to taxes
recommended, would not be costly.  Rather, the challenge would be to
report these impacts in the most understandable and meaningful way. 
For example, the officials said amounts collected could be reported
in a variety of ways, including by fiscal year of the audit closure,
fiscal year of the collection, type of audit, and type of auditor. 


--------------------
\16 IRS also has developed a measure that only tracks collections of
recommended assessments to which taxpayers agreed at the closure of
CEP audits.  Further, IRS has developed a measure of total
collections for all enforcement functions but these collections are
not linked to the taxes recommended. 


   CONCLUSIONS
------------------------------------------------------------ Letter :7

In considering the IRS-wide results of the audit function, analyzing
and reporting the assessment and collection of amounts recommended
provides a more complete picture of revenue impacts than that offered
by looking just at recommended amounts.  For example, our analyses of
audits closed in fiscal year 1992 showed that IRS assessed and
collected a fraction of the recommended additional taxes.  By
disputing recommended taxes, taxpayers substantially reduced the
additional taxes that were ultimately assessed.  IRS, however, did
collect most of the assessed taxes.  A closer analysis showed
differences by the types of audit.  IRS was more likely to assess
taxes recommended in simpler audits of individuals than in complex
audits of corporations.  However, IRS was more likely to collect
assessments from the corporations. 

We believe that analyses of existing IRS data can offer a fuller,
more balanced picture of what happens to additional taxes recommended
in audits.  If IRS and Congress had access to data and analyses on
the assessment and collection of recommended amounts by type of
audit, they would have a more informed foundation for discussions and
decisions on the audit function.  For example, if certain types of
audits recommend taxes that tend not to be assessed or collected, IRS
may decide to analyze the reasons why and then make improvements to
those audits or shift audit resources elsewhere. 

Another broad view of audit impacts would involve ratios that compare
the direct tax revenues generated with IRS' related costs.  Such
ratios can be developed with existing data on the direct costs
incurred to recommend, assess, and collect additional taxes as a
result of the audits.  However, such ratios are not yet complete
measures of audit results.  For example, IRS' data did not include
its direct staff costs to collect the additional taxes and its
indirect costs for the audit, assessment, and collection activity. 
Although incomplete as measures, the ratios provide more information
on audit impacts compared with solely using data on additional taxes
recommended in audits.  These analyses could be enriched if IRS had
data on its direct collection costs and its indirect costs for the
audit, settlement, and collection activities.  IRS is developing an
activity-based costing model that could help IRS to account for these
costs.  The analyses could also benefit from IRS having data on the
other indirect and marginal effects of audits on tax collections and
costs but compiling such data is difficult to do.  We are making no
recommendations on these indirect and marginal effects because we did
not attempt to collect and analyze data on these effects. 


   RECOMMENDATIONS
------------------------------------------------------------ Letter :8

We recommend that the Commissioner of Internal Revenue develop
meaningful ways to report the results to Congress from tracking, over
a reasonable number of future years, existing IRS data on the
assessment and collection of additional amounts recommended in
specific types of audits closed for each fiscal year.  One option for
developing reporting formats could be the tables used in this report. 
The reports would provide fuller measures of the impacts of audits
across IRS than those just on taxes recommended. 

We also recommend that the Commissioner develop a way to track the
direct staff costs of collecting tax assessments associated with
specific types of audits.  Similarly, the Commissioner also should
determine how to account for IRS' indirect costs in auditing returns,
settling audit disputes, and collecting audit assessments by type of
audits.  In analyzing how to account for these indirect costs, IRS
may find that the activity-based costing model being developed can
serve as a helpful tool. 


   AGENCY COMMENTS AND OUR
   EVALUATION
------------------------------------------------------------ Letter :9

On April 27, 1998, we obtained comments on a draft of this report
during a meeting with officials representing IRS.  They included the
Assistant Commissioner for Examination and his staff, the National
Director for Financial Analysis and his staff, the National Director
for Compliance Specialization, and representatives for the National
Director of Appeals, the Assistant Commissioner for Collection, and
the National Director, Legislative Affairs Division.  The IRS
Commissioner also documented the comments in a letter dated May 27,
1998 (see app.  V). 

Both at the meeting and in its letter, IRS agreed to implement our
recommendations.  For our recommendation on reporting the amounts of
recommended taxes that are assessed and collected, IRS said it will
annually report to Congress, by fiscal year, the amounts of
recommended taxes that are collected.  For our recommendation on
tracking direct and indirect IRS costs associated with audits, IRS
said it will continue to develop the activity-based costing model to
track these costs by types of audit. 

IRS' letter included an enclosure that provided various technical
comments on issues discussed in our report as well as other issues. 
These comments dealt with issues such as (1) the need to carefully
analyze and interpret ERIS data; (2) the challenges of allocating
IRS' costs; (3) the value of the activity-based costing model in
allocating costs; (4) the ongoing use of ERIS data; (5) IRS' efforts
since 1992 to improve the audit and dispute resolution processes; (6)
concerns about misinterpretations of the analyses on how much of the
recommended tax amounts were settled, assessed, and collected; and
(7) the need to analyze new measures.  We have made changes and
incorporated those comments that had a direct bearing on the
information provided in this report. 


---------------------------------------------------------- Letter :9.1

We are sending copies of this report to the Chairmen and Ranking
Minority Members of the House Committee on Ways and Means, the Senate
Committee on Finance, and other congressional committees with
responsibility for IRS oversight; the IRS Commissioner; the Director
of the Office of Management and Budget; the Secretary of the
Treasury; and other interested parties.  We will also make the report
available to others upon request. 


Major contributors to this report are listed in appendix VI.  Please
contact me on (202) 512-9110 if you or your staff have any questions
about this report. 

Sincerely yours,

James R.  White
Associate Director, Tax Policy and
Administration Issues


IRS' ENFORCEMENT REVENUE
INFORMATION SYSTEM
=========================================================== Appendix I

The Enforcement Revenue Information System (ERIS) is an automated
data repository, outside the Internal Revenue Service (IRS)
enforcement process, that contains IRS data on its enforcement
results; these data come from several IRS sources.  IRS developed
ERIS to track information on the resolution of enforcement cases. 
Prior to the development of ERIS in 1990, IRS did not have a system
that tracked the enforcement results from each fiscal year. 

The purpose of ERIS is to account for revenues collected and costs
incurred as a result of IRS enforcement activities.  In addition,
ERIS provides a link to taxes assessed and collected for different
types of cases tracked by enforcement activities.  ERIS enhances IRS'
ability to provide a more complete reporting of enforcement results
and forecast enforcement revenues.  ERIS also may help IRS manage its
enforcement activities better to the extent that it provides more
complete reporting of costs and revenues. 

ERIS works by integrating data from the various enforcement functions
with corresponding Master File data to build a comprehensive
enforcement database.  It merges data extracted from the Audit
Information Management System, Information Reporting Program Case
Analysis System, Individual and Business Master Files, Individual
Retirement Account Master File, and Non-Master File.  Once the data
are integrated from various sources, IRS develops a summary database
from which comprehensive reports are printed. 


STATUS OF ADDITIONAL AMOUNTS
RECOMMENDED FOR AUDITS CLOSED IN
FISCAL YEARS 1993-1997 ACROSS
SEVEN TYPES OF AUDITS
========================================================== Appendix II



                                    Table II.1
                     
                     Status of Additional Amounts Recommended
                      for Audits Closed in Fiscal Year 1993
                       Across Seven Types of Audits, as of
                                September 27, 1997

                              (Dollars in millions)

                     Settlement status of
                     recommended amounts\a              Collection status
                -------------------------------  --------------------------------
Ty
pe
of                                                                     As percent
au              Percent of    Percent            Collecte  As percent   of amount
di  Recommende    disputes        not   Percent         d   of amount  recommende
t    d amounts   unsettled   assessed  assessed   amounts    assessed           d
--  ----------  ----------  ---------  --------  --------  ----------  ----------
Se      $1,146          0%        23%       77%      $438         50%          38
 r
 v
 i
 c
 e
 c
 e
 n
 t
 e
 r
In       2,721           6         31        64       947          55          35
 d
 i
 v
 i
 d
 u
 a
 l
 n
 o
 n
 b
 u
 s
 i
 n
 e
 s
 s
In       1,331           9         32        59       363          46          27
 d
 i
 v
 i
 d
 u
 a
 l
 b
 u
 s
 i
 n
 e
 s
 s
Sm       1,292           5         45        50       300          46          23
 a
 l
 l
 c
 o
 r
 p
 o
 r
 a
 t
 i
 o
 n
La       2,821          13         53        34       623          64          22
 r
 g
 e
 c
 o
 r
 p
 o
 r
 a
 t
 i
 o
 n
Co      10,547          43         29        28     2,744          93          26
 o
 r
 d
 i
 n
 a
 t
 e
 d
 E
 x
 a
 m
 P
 r
 o
 g
 r
 a
 m
Ot       2,164           7         51        42       606          67          28
 h
 e
 r
 \
 b
=================================================================================
19     $22,022          25         35        40    $6,022          68          27
 93
 T
 o
 t
 a
 l
 \
 c
---------------------------------------------------------------------------------
\a Percentages do not total to 100 percent due to rounding. 

\b "Other" includes audits of returns for employment tax, estate tax,
excise tax, and gift tax; audits conducted in IRS training; and
audits categorized by IRS as other. 

\c Amounts do not add to total due to rounding. 

Source:  GAO analysis of IRS ERIS data. 



                                    Table II.2
                     
                     Status of Additional Amounts Recommended
                      for Audits Closed in Fiscal Year 1994
                       Across Seven Types of Audits, as of
                                September 27, 1997

                              (Dollars in millions)

                      Settlement status of
                      recommended amounts              Collection status
                  ----------------------------  --------------------------------
                   Percent
Type                    of                                            As percent
of                disputes   Percent            Collecte  As percent   of amount
audi  Recommende  unsettle       not   Percent         d   of amount  recommende
t      d amounts         d  assessed  assessed   amounts    assessed           d
----  ----------  --------  --------  --------  --------  ----------  ----------
Serv        $830        0%       18%       82%      $353         52%         43%
 ice
 cen
 ter
Indi       3,330         7        23        70       965          41          29
 vid
 ual
 non
 bus
 ine
 ss
Indi       1,428         8        25        67       364          38          26
 vid
 ual
 bus
 ine
 ss
Smal         868         6        37        57       247          50          28
 l
 cor
 por
 ati
 on
Larg       1,867        10        40        50       622          67          33
 e
 cor
 por
 ati
 on
Coor      12,228        48        26        26     2,994          95          24
 din
 ate
 d
 Exa
 m
 Pro
 gra
 m
Othe       2,087        31        35        34       514          72          25
 r\a
================================================================================
1994     $22,637        31        28        41    $6,059          65          27
 Tot
 al\
 b
--------------------------------------------------------------------------------
\a "Other" includes audits of returns for employment tax, estate tax,
excise tax, and gift tax; audits conducted in IRS training; and
audits categorized by IRS as other. 

\b Amounts do not add to total due to rounding. 

Source:  GAO analysis of IRS ERIS data. 



                                    Table II.3
                     
                     Status of Additional Amounts Recommended
                      for Audits Closed in Fiscal Year 1995
                       Across Seven Types of Audits, as of
                                September 27, 1997

                              (Dollars in millions)

                      Settlement status of
                     recommended amounts\a             Collection status
                  ----------------------------  --------------------------------
                   Percent
Type                    of                                            As percent
of                disputes   Percent            Collecte  As percent   of amount
audi  Recommende  unsettle       not   Percent         d   of amount  recommende
t      d amounts         d  assessed  assessed   amounts    assessed           d
----  ----------  --------  --------  --------  --------  ----------  ----------
Serv      $1,300        1%       16%       83%      $409         38%         31%
 ice
 cen
 ter
Indi       3,599        11        23        66       796          34          22
 vid
 ual
 non
 bus
 ine
 ss
Indi       1,730         5        24        70       355          29          21
 vid
 ual
 bus
 ine
 ss
Smal         905        24        33        43       220          56          24
 l
 cor
 por
 ati
 on
Larg       2,460        40        28        32       661          84          27
 e
 cor
 por
 ati
 on
Coor      15,406        59        14        27     3,792          91          25
 din
 ate
 d
 Exa
 m
 Pro
 gra
 m
Othe       1,783        16        32        52       644          70          36
 r\b
================================================================================
1995     $27,184        41        19        40    $6,878          63          25
 Tot
 al\
 c
--------------------------------------------------------------------------------
\a Percentages do not total to 100 percent due to rounding. 

\b "Other" includes audits of returns for employment tax, estate tax,
excise tax, and gift tax; audits conducted in IRS training; and
audits categorized by IRS as other. 

\c Amounts do not add to total due to rounding. 

Source:  GAO analysis of IRS ERIS data. 



                                    Table II.4
                     
                     Status of Additional Amounts Recommended
                      for Audits Closed in Fiscal Year 1996
                       Across Seven Types of Audits, as of
                                September 27, 1997

                              (Dollars in millions)

                      Settlement status of
                     recommended amounts\a             Collection status
                  ----------------------------  --------------------------------
                   Percent
Type                    of                                            As percent
of                disputes   Percent            Collecte  As percent   of amount
audi  Recommende  unsettle       not   Percent         d   of amount  recommende
t      d amounts         d  assessed  assessed   amounts    assessed           d
----  ----------  --------  --------  --------  --------  ----------  ----------
Serv      $1,468        1%       16%       83%      $469         38%         32%
 ice
 cen
 ter
Indi       3,379        17        18        65       826          38          24
 vid
 ual
 non
 bus
 ine
 ss
Indi       1,633        15        16        69       363          32          22
 vid
 ual
 bus
 ine
 ss
Smal         880        25        27        48       251          60          29
 l
 cor
 por
 ati
 on
Larg       3,135        42        16        41       762          59          24
 e
 cor
 por
 ati
 on
Coor      18,093        63        17        20     3,331          92          18
 din
 ate
 d
 Exa
 m
 Pro
 gra
 m
Othe       2,183        16        34        50       734          67          34
 r\b
================================================================================
1996     $30,771        46        18        36    $6,737          61          22
 Tot
 al\
 c
--------------------------------------------------------------------------------
\a Percentages do not total to 100 percent due to rounding. 

\b "Other" includes audits of returns for employment tax, estate tax,
excise tax, and gift tax; audits conducted in IRS training; and
audits categorized by IRS as other. 

\c Amounts do not add to total due to rounding. 

Source:  GAO analysis of IRS ERIS data. 



                                    Table II.5
                     
                     Status of Additional Amounts Recommended
                      for Audits Closed in Fiscal Year 1997
                       Across Seven Types of Audits, as of
                                September 27, 1997

                              (Dollars in millions)

                      Settlement status of
                     recommended amounts\a             Collection status
                  ----------------------------  --------------------------------
                   Percent
Type                    of                                            As percent
of                disputes   Percent            Collecte  As percent   of amount
audi  Recommende  unsettle       not   Percent         d   of amount  recommende
t      d amounts         d  assessed  assessed   amounts    assessed           d
----  ----------  --------  --------  --------  --------  ----------  ----------
Serv      $2,343        2%        6%       92%      $328         15%         14%
 ice
 cen
 ter
Indi       3,544        34         7        60       688          33          19
 vid
 ual
 non
 bus
 ine
 ss
Indi       2,107        33         9        58       309          25          15
 vid
 ual
 bus
 ine
 ss
Smal       1,109        43        10        46       290          56          26
 l
 cor
 por
 ati
 on
Larg       3,158        54        10        36       712          63          23
 e
 cor
 por
 ati
 on
Coor      16,445        70         7        23     3,512          93          21
 din
 ate
 d
 Exa
 m
 Pro
 gra
 m
Othe       2,984        56         6        38       608          53          20
 r\b
================================================================================
1997     $31,691        55         7        38    $6,447          54          20
 Tot
 al\
 c
--------------------------------------------------------------------------------
\a Percentages do not total to 100 percent due to rounding. 

\b "Other" includes audits of returns for employment tax, estate tax,
excise tax, and gift tax; audits conducted in IRS training; and
audits categorized by IRS as other. 

\c Amounts do not add to total due to rounding. 

Source:  GAO analysis of IRS ERIS data. 


DETAILED INFORMATION ON AUDIT
RESULTS
========================================================= Appendix III



                                   Table III.1
                     
                     Amounts Recommended for Audits Closed in
                         Fiscal Year 1992 Across 30 Audit
                     Subcategories, as of September 27, 1997

                              (Dollars in millions)

                           Settlement status\a           Collection status
                        --------------------------  ----------------------------
                         Percent                                  As
                              of            Percen  Collec   percent  As percent
                 Total  disputes   Percent       t     ted        of   of amount
              recommen  unsettle       not  assess  amount    amount  recommende
Subcategory        ded         d  assessed      ed       s  assessed           d
------------  --------  --------  --------  ------  ------  --------  ----------
Service center
--------------------------------------------------------------------------------
Individual    $1,310.8        0%       24%     76%  $542.0       54%         41%
 nonbusiness         5                                   6
Individual      107.98         1        23      76   62.54        76          58
 business
Other            44.80         0        33      67   20.69        69          46

Individual nonbusiness
--------------------------------------------------------------------------------
1040A TPI <     373.27         0        21      79  126.41        43          34
 $25,000
Non-1040A       237.66         1        25      74   77.77        44          33
 TPI <
 $25,000
TPI $25,000     375.23         1        26      73  135.96        50          36
 < $50,000
TPI $50,000     382.25         2        32      66  156.54        62          41
 < $100,000
TPI $100,000  1,394.72         7        48      46  397.13        62          28
 and over

Individual business
--------------------------------------------------------------------------------
C-TGR <         123.00         0        19      81   30.84        31          25
 $25,000
C-TGR           317.51         1        30      70   84.51        38          27
 $25,000 <
 $100,000
C-TGR           760.07         8        38      54  220.21        53          29
 $100,000
 and over
F-TGR <          14.41         0        29      71    7.95        78          55
 $100,000
F-TGR            75.38        23        30      47   27.77        79          37
 $100,000
 and over

Small corporations
--------------------------------------------------------------------------------
No balance       72.16         5        42      53   16.53        44          23
 sheet
Assets <        167.76         6        40      55   38.89        43          23
 $250,000
Assets          199.65         4        41      55   67.43        61          34
 $250,000 <
 $1 million
Assets $1       363.77         7        41      52  123.71        65          34
 million <
 $5 million
Assets $5       197.26         8        46      47   54.56        59          28
 million <
 $10 million

Large corporations
--------------------------------------------------------------------------------
Assets $10      535.22        10        52      39  146.08        71          27
 million <
 $50 million
Assets $50      288.20        12        47      41   87.67        74          30
 million <
 $100
 million
Assets $100     420.94         7        46      47  107.62        55          26
 million <
 $250
 million
Assets $250     862.43         5        72      24  188.59        93          22
 million and
 over
Foreign         113.80         0        85      15   13.06        79          11
 corporations
Coordinated   13,893.3        39        41      20  2,757.        97          20
 Exam                2                                  19
 Program

Other returns
--------------------------------------------------------------------------------
Employment      287.03         1        52      48   71.95        53          25
 tax
Estate tax    1,161.84        43        28      30  308.84        90          27
Excise tax      213.03         2        39      59   48.98        39          23
Gift tax        172.93        27        51      22   36.44        96          21
Trainee         328.34        11        24      65  166.84        78          51
Other            38.57         3        52      45   12.77        73          33
================================================================================
Total         $24,833.        26        40      34  $6,137        72          25
                    38                                 .54
--------------------------------------------------------------------------------
Legend

Assessed = Tax and penalty assessed less related abatements
1040A = Nonbusiness returns filed by individuals
TPI = Total positive income (income reported as a positive on tax
return tables)
C-TGR = Form 1040 Schedule C (profit or loss from business) total
gross receipts
F-TGR = Form 1040 Schedule F (profit or loss from farming) total
gross receipts

\a Percentages do not total to 100 percent due to rounding. 

Source:  GAO analysis of IRS ERIS data. 



                Table III.2 Taxes Recommended, Assessed,
                and Collected per Direct Staff Cost for
                Audits Closed in Fiscal Year 1992 Across
                   the 30 Audit Subcategories, as of
                           September 27, 1997


                                   Recommended    Assessed  Collection
                                       to cost     to cost     to cost
Subcategory                            ratio\a     ratio\b     ratio\c
---------------------------------  -----------  ----------  ----------
Service center
----------------------------------------------------------------------
Individual nonbusiness                   216:1       112:1        61:1
Individual business                      120:1        66:1        51:1
Other                                     93:1        58:1        40:1

Individual nonbusiness
----------------------------------------------------------------------
1040A TPI < $25,000                       32:1        22:1         9:1
Non-1040A TPI < $25,000                   19:1        12:1         5:1
TPI $25,000 < $50,000                     19:1        11:1         6:1
TPI $50,000 < $100,000                    20:1        11:1         7:1
TPI $100,000 and over                     44:1        16:1        10:1

Individual business re
----------------------------------------------------------------------
C-TGR < $25,000                           20:1        15:1         5:1
C-TGR $25,000 < $100,000                  19:1        12:1         4:1
C-TGR $100,000 and over                   22:1        11:1         6:1
F-TGR < $100,000                          10:1         6:1         5:1
F-TGR $100,000 and over                   23:1        10:1         8:1

Small corporation retu
----------------------------------------------------------------------
No balance sheet                          25:1        12:1         5:1
Assets < $250,000                         11:1         5:1         2:1
Assets $250,000 < $1,000,000              11:1         6:1         4:1
Assets $1,000,000 < $5,000,000            16:1         8:1         5:1
Assets $5,000,000 < $10,000,000           22:1         9:1         6:1

Large corporations
----------------------------------------------------------------------
Assets $10,000,000 < $50,000,000          26:1         9:1         7:1
Assets $50,000,000 < $100,000,000         36:1        14:1        10:1
Assets $100,000,000 <                     46:1        19:1        10:1
 $250,000,000
Assets $250,000,000 and over              74:1        15:1        14:1
Foreign corporations                     108:1        12:1         9:1
Coordinated Exam Program                 124:1        24:1        24:1

Other returns
----------------------------------------------------------------------
Employment tax                            31:1        13:1         7:1
Estate tax                                67:1        18:1        16:1
Excise tax                                30:1        17:1         6:1
Gift tax                                 120:1        22:1        21:1
Trainee                                    9:1         6:1         4:1
Other                                      1:1         1:1           0
======================================================================
Total                                     50:1        16:1        11:1
----------------------------------------------------------------------
Legend

1040A = Nonbusiness returns filed by individuals
TPI = Total positive income (income reported as a positive on tax
return tables)
C-TGR = Form 1040 Schedule C (profit or loss from business) total
gross receipts
F-TGR = Form 1040 Schedule F (profit or loss from farming) total
gross receipts

\a Includes Examination direct staff cost but excludes indirect cost. 

\b Includes Examination, Appeals, and Counsel direct staff cost but
excludes indirect cost. 

\c Includes Examination, Appeals, and Counsel staff cost but excludes
Collection's direct cost and indirect cost. 

Source:  GAO analysis of IRS ERIS data. 


THE AMOUNT OF TIME IRS TAKES TO
COLLECT ADDITIONAL TAXES ASSESSED
AFTER AUDITS
========================================================== Appendix IV

In developing ratios of the amount of taxes collected to the direct
costs for the audit, assessment, and collection activities, we could
not include IRS' direct staff costs for the collection activity.  We
tried various analyses to gain insights on these costs to collect
audit-based assessments, but none of our analyses were conclusive. 
In sum, IRS did not have data that would indicate the significance of
these costs. 

For example, we found that about 10 percent of the additional taxes
collected were collected through the direct involvement of Collection
staff.  However, IRS' data did not help us to translate this
information into the related direct staff costs.  Nor did IRS have
enough data to allow us to develop formulas for allocating the
Collection Division's overall staff costs to the direct staff costs
of collecting audit-based tax assessments. 

Although our analyses did not help us compute the direct collection
costs, we are reporting our results on how long IRS took to collect
the audit-based tax assessments.  To determine how much time IRS took
to collect the tax assessments associated with audits closed in
fiscal year 1992, we analyzed ERIS data on the amount of taxes
collected from the various types of collection notices.  Using IRS
manuals, we determined the number of weeks that was to have elapsed
between the assessment and each notice.\1 These collections also
accounted for amounts that taxpayers paid as a result of the audit
before IRS made the official assessment.  We analyzed the timing of
collections by types of audits. 

For audits closed in fiscal year 1992, IRS had collected $6.1 billion
of the $8.5 billion in additional assessments, as of September 27,
1997.\2 Although the time taken to do the audits and settle any
disputes can be quite lengthy, IRS usually collected any additional
taxes from doing the audits prior to or soon after assessment; about
81 percent was collected prior to assessment and within the first 5
weeks after assessment.  IRS collects taxes prior to assessment to
the extent that taxpayers overwithhold their income taxes,
overestimate their quarterly tax payments for the audited tax return,
carry over excess tax payments from previous tax returns, or make a
payment prior to the additional assessment to prevent the accrual of
further interest. 

Specifically, these analyses showed that IRS collected taxes sooner
from corporations than from individuals.  IRS collected 95 percent of
the taxes from CEP corporations and 92 percent from other large
corporations before the assessments or within the first 5 weeks of
the assessments.  For individuals audited at service centers or
district offices, IRS collected about 60 percent of the taxes within
these time periods.  However, the portion of the taxes that IRS
collected after 15 weeks was much higher for individuals than for
corporations; staff from the Collection Division become involved
after this time.  For example, less than 5 percent of the taxes
collected from CEP and other large corporations were collected after
15 weeks.  About 25 percent of the taxes collected from individuals
audited at the service centers and district offices were collected
after this period. 

Table IV.1 and table IV.2 show the timing of collections of assessed
taxes for the various types of audits closed in fiscal year 1992. 



                                    Table IV.1
                     
                      Timing of Collection of Assessed Taxes
                     From Audits Closed in Fiscal Year 1992,
                            as of September 27, 1997\a

                              (Dollars in millions)

                                                          Percent
                                           Percent      collected        Percent
                            Percent      collected   between 5 to      collected
Type                      collected       within 5       15 weeks  over 15 weeks
of            Amount       prior to    weeks after          after          after
audit      collected     assessment     assessment     assessment     assessment
-----  -------------  -------------  -------------  -------------  -------------
Servi           $625            27%            32%            17%            25%
 ce
 cent
 er
Indiv            894             37             23             16             24
 idual
 nonb
 usin
 ess
Indiv            371             36             23             13             28
 idual
 busi
 ness
Small            301             54             30              8              9
 corp
 orat
 ions
Large            543             59             33              5              4
 corp
 orat
 ions
Coord          2,757             62             33              3              1
 inat
 ed
 Exam
 inat
 ion
 Prog
 ram
Other            646             60             18              9             12
 \b
================================================================================
Total         $6,138             52             29              8             10
 \c
--------------------------------------------------------------------------------
\a Percentages do not total to 100 percent due to rounding. 

\b "Other" includes audits of returns for employment tax, estate tax,
excise tax, and gift tax; audits conducted in IRS training; and
audits categorized by IRS as other. 

\c Amounts do not add to total due to rounding. 

Source:  GAO analysis of IRS ERIS data. 



                                    Table IV.2
                     
                      Timing of Collection of Assessed Taxes
                      From Audits Closed in Fiscal Year 1992
                     Across the 30 Audit Subcategories, as of
                               September 27, 1997\a

                              (Dollars in millions)

                                                           Percent
                                             Percent     collected       Percent
                               Percent     collected  between 5 to     collected
                             collected      within 5      15 weeks       over 15
Subcategor           Tax      prior to   weeks after         after   weeks after
y              collected    assessment    assessment    assessment    assessment
----------  ------------  ------------  ------------  ------------  ------------
Service center
--------------------------------------------------------------------------------
Individual       $542.06           27%           30%           17%           26%
 nonbusine
 ss
Individual         62.54            32            33            17            17
 business
Other              20.69            14            72             6             8

Individual nonbusiness
--------------------------------------------------------------------------------
1040A TPI         126.41            11            19            29            41
 < $25,000
Non-1040A          77.77            22            20            22            36
 TPI <
 $25,000
TPI               135.96            28            21            18            32
 $25,000 <
 $50,000
TPI               156.54            41            21            14            24
 $50,000 <
 $100,000
TPI               397.13            50            26            10            14
 $100,000
 and over

Individual business returns
--------------------------------------------------------------------------------
C-TGR <            30.84            20            17            19            45
 $25,000
C-TGR              84.51            26            19            15            40
 $25,000
 <
 $100,000
C-TGR             220.21            39            24            12            24
 $100,000
 and over
F-TGR <             7.95            61            14             8            17
 $100,000
F-TGR              27.77            48            34             8            10
 $100,000
 and over

Small corporation returns
--------------------------------------------------------------------------------
No balance         16.53            58            31             3             7
 sheet
Assets <           38.89            51            24            10            16
 $250,000
Assets             67.43            52            28            10            10
 $250,000
 < $1
 million
Assets $1         123.71            56            30             7             7
 million
 < $5
 million
Assets $5          54.56            52            33             6             9
 million
 < $10
 million

Large corporation returns
--------------------------------------------------------------------------------
Assets $10        146.08            55            34             4             7
 million
 < $50
 million
Assets $50         87.67            53            35             9             4
 million
 < $100
 million
Assets            107.62            65            26             8             2
 $100
 million
 < $250
 million
Assets            188.59            59            37             3             2
 $250
 million
 and over
Foreign            13.06            93             6             0             1
 corporati
 ons
Coordinate      2,757.19            62            33             3             1
 d Exam
 Program

Other returns
--------------------------------------------------------------------------------
Employment         71.95            40            25            12            23
 tax
Estate tax        308.84            80            13             4             3
Excise tax         48.98            52            21            10            18
Gift tax           36.44            69            20             7             3
Trainee           166.84            34            22            18            26
Other              12.77            62            27             5             6
================================================================================
Total\b        $6,137.54            52            29             8            10
--------------------------------------------------------------------------------
Legend

1040A = Nonbusiness returns filed by individuals
TPI = Total positive income (income reported as a positive on tax
return tables)
C-TGR = Form 1040 Schedule C (profit or loss from business) total
gross receipts
F-TGR = Form 1040 Schedule F (profit or loss from farming) total
gross receipts

\a Percentages do not total to 100 percent due to rounding

\b Amounts do not add to total due to rounding. 

Source:  GAO analysis of IRS ERIS data. 



(See figure in printed edition.)Appendix V

--------------------
\1 IRS sends up to four notices at 5-week intervals to individual
taxpayers.  To business taxpayers, IRS is to send up to three
notices, with the second notice being sent 5 weeks after the first
and the third notice being sent 6 weeks after the second. 

\2 We did not analyze how long the remaining $2.4 billion had gone
uncollected; nor did we analyze the length of the audit and
settlement activities, given our focus on the collection of
additional recommended taxes. 


COMMENTS FROM THE INTERNAL REVENUE
SERVICE
========================================================== Appendix IV


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix VI

GENERAL GOVERNMENT DIVISION,
WASHINGTON, D.C. 

Thomas D.  Short, Assistant Director, Tax Policy and Administration
Issues
Kevin E.  Daly, Senior Economist

KANSAS CITY FIELD OFFICE

Royce L.  Baker, Tax Issue Area Coordinator
James A.  Slaterbeck, Evaluator-in-Charge
Bradley L.  Terry, Evaluator
Thomas N.  Bloom, Computer Specialist

*** End of document. ***