Tax Administration: Audit Trends and Taxes Assessed on Large Corporations
(Letter Report, 10/13/95, GAO/GGD-96-6).

GAO reviewed the results of the Internal Revenue Service's (IRS) efforts
to audit the tax returns of about 45,000 large corporations, focusing
on: (1) audit trends for fiscal years 1988 through 1994; (2) the portion
of taxes recommended by agents that were eventually assessed; and (3)
the profiles of audited large corporations compared with those of
nonaudited corporations.

GAO found that: (1) for every dollar invested in large corporation
audits, IRS ultimately assessed $15 in additional taxes for the years
1988 through 1994; (2) IRS invested more hours in directly auditing
large corporations but recommended less additional tax per hour invested
in 1994 compared to 1988; (3) in 1994, large corporations appealed 66
percent of the additional taxes that IRS recommended in its audits; (4)
between 1988 and 1994, IRS assessed 27 percent of the recommended
additional taxes either after agreement or resolution in appeals; (5)
IRS believed that the assessment rate was not an accurate measure of
audit effectiveness, since various factors outside the audit could lower
the rate; (6) the assessment rates ranged from 20 to 38 percent for four
asset classes and from 0 to 103 percent by IRS district, but the reasons
for the disparities were unclear; (7) the rates for audits closing
without any adjustments has rapidly increased, raising questions about
how IRS selects returns for audits; and (8) audited corporations tended
to report higher average incomes, tax liabilities, and other tax amounts
than nonaudited corporations.

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

 REPORTNUM:  GGD-96-6
     TITLE:  Tax Administration: Audit Trends and Taxes Assessed on 
             Large Corporations
      DATE:  10/13/95
   SUBJECT:  Tax administration
             Corporations
             Federal taxes
             Income taxes
             Compliance
             Government collections
             Tax nonpayment
             Tax return audits
             Appellate procedure
             Auditing procedures
IDENTIFIER:  IRS Coordinated Examination Program
             IRS Audit Information Management System
             IRS Business Master File
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* No attempt has been made to display graphic images, although figure    *
* captions are reproduced. Tables are included, but may not resemble     *
* those in the printed version.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


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


Report to the Commissioner, Internal Revenue Service

October 1995

TAX ADMINISTRATION - AUDIT TRENDS
AND TAXES ASSESSED ON LARGE
CORPORATIONS

GAO/GGD-96-6

Large Corporate Audit Data

(268634)


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

  AIMS - Audit Information Management System
  BMF - Business Master File
  CEP - Coordinated Examination Program
  IRS - Internal Revenue Service
  SOI - Statistics of Income
  TIN - taxpayer identification number

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


B-260136

October 13, 1995

The Honorable Margaret Milner Richardson
Commissioner of Internal Revenue

Dear Ms.  Richardson: 

This report focuses on the results of the Internal Revenue Service's
(IRS) program to audit the tax returns of about 45,000 large
corporations.  IRS audits of returns filed by these 45,000
corporations plus the 1,700 largest corporations in IRS' Coordinated
Examination Program (CEP) have generated about two-thirds of the
additional taxes recommended from all income tax audits.  Although we
have reported on CEP,\1 under which IRS audits the largest
corporations, we have not previously studied audits of other large
corporations, particularly trends on what IRS invests in and produces
from these audits.  Until recently, IRS had not analyzed these trends
either. 

We used IRS data to (1) analyze audit trends for fiscal years 1988
through 1994, (2) compute the assessment rate--the portion of taxes
recommended by revenue agents that were eventually assessed, and (3)
develop and compare profiles of audited large corporations with those
not audited.  The second phase of our work, now under way, will
analyze factors that affect the assessment rate and review IRS'
methodology for estimating it. 


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


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

For audit purposes, IRS' Examination Division defines a corporation
as large or small depending on the amount of assets reported on its
income tax return.  Small corporations (about 2.4 million) are
defined as those that report total assets of less than $10 million. 

Corporations reporting higher assets are considered to be large, and
IRS audits large corporations in two groups.  IRS annually selects
about 1,700 of the largest and most complex corporations for CEP. 
The remaining large corporations (about 45,000) may be audited under
a separate IRS program--the subject of this report. 

IRS has different ways to select corporations for audits.  For small
corporations, IRS uses a formula that measures the likelihood of
changes to tax liability.  This formula helps IRS objectively select
returns for audits that are considered to be most likely to produce
tax changes.  IRS developed the formula by analyzing results of
line-by-line audits of a random sample of tax returns.  Once
selected, a small corporation return is usually audited by an IRS
revenue agent.  In fiscal year 1994, IRS audited about 44,000 (1.8
percent) of 2.4 million income tax returns filed by small
corporations. 

For CEP, IRS selects corporations on the basis of criteria for size,
complexity, and the like.  After considering its audit resources and
manually reviewing the audit potential of every CEP return, IRS
selects returns for audit.  IRS audits CEP returns with teams of
revenue agents and specialists, such as economists and engineers. 
Our 1992 report on CEP noted that IRS audited about 77 percent of the
CEP returns for fiscal year 1991.\2

The remaining large corporations (hereafter referred to as large
corporations) are usually selected for audit on the basis of IRS
agents' judgment, rather than through a scoring formula or specific
criteria.  In some cases, revenue agents at a service center select
returns and send them to a district office to be audited; in other
cases, all relevant returns are sent to the district office, where
revenue agents choose those they will audit.  Unlike with CEP
returns, IRS usually uses a revenue agent (hereafter referred to as
auditor) rather than a team to audit the returns from this segment of
the large corporation universe.  According to IRS Examination
officials, these individual auditors recently have been using IRS
specialists more than they have in the past. 

To give perspective on the sizes of these large corporations compared
to other types of businesses, we analyzed average assets reported for
1992 (the most recent year of available data at IRS' Statistics of
Income Division).  Reported assets ranged from an average of about
$0.4 million by small corporations to about $6.8 billion by CEP
corporations.  Within this wide range, the remaining large
corporations reported an average of $130.7 million in assets as
compared with partnerships, which reported an average of $1.3 million
in assets. 

In recording the audit results for large corporations not part of
CEP, IRS has created four classes according to asset size, ranging
from $10 million to over $250 million.  To facilitate our reporting
of trends, we collapsed the four classes into two, (1) lower asset
($10 million to less than $100 million) and (2) higher asset ($100
million and over).  Narrative in this letter and appendix II focuses
on the differences in the trends for the two combined classes but
also discusses the four classes, particularly their assessment rates. 


--------------------
\2 Tax Administration:  IRS Efforts to Improve Corporate Compliance
(GAO/GGD-92-81BR, April 17, 1992). 


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

IRS' return on the large corporation audit program appears to be
high.  For each dollar invested directly in audits, IRS recommended
$56 and ultimately assessed $15 in additional taxes for the years
1988 through 1994.  These calculations exclude indirect audit costs
(e.g., overhead); other IRS costs (e.g., appeals, litigation); and
corporations' costs.  They also exclude direct audit costs associated
with recommended taxes for which the assessed amount has not been
finalized. 

IRS invested more in directly auditing large corporations but
recommended less additional tax per hour invested in 1994 compared to
1988.  IRS spent 25 percent more hours and audited only 3 percent
more returns.  Further, additional taxes recommended per audit hour
decreased by 7 percent in current dollars.  In 1994 constant dollars,
additional taxes recommended decreased 4 percent overall, decreased 7
percent per return audited, and decreased 23 percent per audit hour. 
IRS Examination officials explained that the increased hours stemmed
from auditing more complex returns and issues.  As a result, IRS
specialists also spent more time on the audits.  These officials
noted that the tax yield from making these investments lagged because
of the complexity, among other reasons. 

If IRS auditors conclude that additional taxes are owed, taxpayers
may appeal within IRS or agree to pay.  IRS assesses any amounts the
taxpayer agrees to pay and any appealed amounts ultimately resolved
as taxes owed.  Between 1988 and 1994, large corporations agreed with
ever-higher portions of recommended amounts but still agreed with
only 34 percent of the amounts by 1994.  They continued to appeal
most recommended amounts--they appealed 66 percent for 1994.  IRS
appeared to lose most appealed amounts.  Of the additional taxes
auditors recommended over the 7 years, IRS assessed 27 percent after
either agreement or resolution in appeals.  IRS Examination officials
noted that various factors outside the audit (e.g., economic
conditions, claims for refunds, or net operating losses) could lower
the assessment rate.  Thus, they cautioned against using the rate as
a measure of audit effectiveness. 

This assessment rate varied widely when disaggregated.  By the four
asset classes, the rates ranged from 20 percent to 38 percent.  By
IRS District, the rate reached as high as 103 percent and fell to
less than 1 percent at 20 IRS districts that recommended over $100
million in additional taxes during the 7 years we analyzed.  The
reasons for such disparate rates were not apparent, but IRS
Examination officials said they were attempting to find out.\3

IRS auditors can close audits without recommending additional tax
changes.  Audits that end with no change to taxes owed could have
adjustments (e.g., reduced a reported net operating loss but not
enough to produce a tax liability) or no adjustments.  IRS views the
former as productive and the latter as unproductive.  These trends
are moving in the wrong direction, from what IRS expects.  Audits
that adjusted taxable income but not enough to change taxes owed
dropped over 40 percent.  Such adjustments can reduce the net
operating loss for the audited year, which also reduces carryover of
that loss to other years.  The rates for audits closing without any
adjustments at all almost doubled. 

These no-change trends raise questions about how well IRS selects
returns for audit and/or audits them--questions we plan to explore in
the second phase of our work.  IRS has also been concerned about
these and other trends.  IRS Examination officials said they convened
task forces in 1994 to find ways to select better returns for audit
and to then better focus the audits. 

Our profile of these large corporations showed that most were engaged
in manufacturing or the finance/insurance industries in 1992. 
Audited corporations tended to report higher average income, tax
liability, and other tax amounts than nonaudited corporations. 


--------------------
\3 Rates can exceed 100 percent for such reasons as Appeals assessing
more taxes than IRS auditors recommended.  Rates can drop
significantly whenever the recommended taxes are, among other
reasons, reduced or offset by claims filed by the large corporations. 


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

Our objectives were to (1) analyze audit trends in fiscal years 1988
through 1994 for large corporations, (2) compute their assessment
rate, and (3) develop and compare profiles of audited large
corporations with those not audited.  To identify large corporations,
we used IRS data on those reporting assets of $10 million and more. 
We used IRS data to eliminate CEP corporations. 

We used three IRS databases to meet our objectives.  To analyze audit
trends, we used Audit Information Management System (AIMS) data on
large corporate audits closed in fiscal years 1988 through 1994.  To
compute the assessment rate, we computer-matched the AIMS data on
recommended tax assessments to actual tax assessments on the Business
Master File (BMF), which contains information about business tax
returns.  We tracked the BMF data through December 1994.  To develop
a profile of the large corporations, we obtained the 1992 Statistics
of Income (SOI) file for corporations--the most recent at the time we
did our work.  We matched the SOI and AIMS data to divide our
population into audited and not audited groups. 

We asked IRS Examination officials at the National Office to review
our analyses of the audit trends and assessment rates and to provide
any explanations.  We performed our work in Washington, D.C., and
Mission, KS, between May 1994 and May 1995 in accordance with
generally accepted government auditing standards.  Appendix I has
more information on our objectives, scope, and methodology. 


   AUDIT TRENDS
------------------------------------------------------------ Letter :4

Figure 1 summarizes trends in large corporate audits for fiscal years
1988 through 1994. 

   Figure 1:  Changes in Measures
   of Large Corporation Audits
   Between Fiscal Years 1988 and
   1994

   (See figure in printed
   edition.)

\a This represents IRS' expected direction of these indicators for
the period 1988 to 1994.  This is not intended to reflect IRS' longer
term goals. 

\b Presented in 1994 constant dollars. 

Source:  GAO analysis of IRS data. 

More details on these trends follow. 

1.  Number of audits:  The total number of audited returns increased
about 3 percent (from 10,062 in 1988 to 10,392 in 1994), after
peaking in 1991 at 11,962.  This increase largely involved
corporations with less than $50 million in assets.  The number of
audited returns for higher asset corporations fluctuated but
decreased 16 percent between 1988 and 1994, particularly in 1989 and
1990.  (Refer to table II.1.)

2.  Audit coverage:  Audit coverage rose from 23 percent in 1988 to
24 percent in 1994, after peaking at 28 percent in 1991.  Coverage
varied by asset class.  It decreased (especially from 1991 to 1992)
from 43 percent to 31 percent for higher asset corporations, and it
increased from 18 percent to 22 percent for those with lower assets
after peaking in 1991 at 25 percent.  (Refer to table II.2.)

3.  Direct audit hours:  A comparison of 1988 and 1994 shows that IRS
invested 25 percent more hours in auditing large corporations,
particularly those with lower assets.  Audit hours decreased for
higher asset corporations, particularly 1988 through 1990.  (Refer to
table II.3.)

4.  Direct audit hours per return:  This ratio increased 21 percent
from 1988 to 1994, driven by audits of lower asset corporations. 
Their ratio increased 43 percent compared to 14 percent for higher
asset corporations.  On average, IRS spent twice as long auditing a
return with higher assets compared to those with lower assets--184
hours versus 89 hours, respectively.  (Refer to table II.4.)

5.  Additional recommended taxes:  This amount increased 16 percent
from about $1.6 billion in 1988 to $1.9 billion in 1994.  It also
increased for both types of corporations.  The amount peaked in 1992
for lower asset corporations and fluctuated year to year for higher
asset corporations while peaking in 1993.  IRS Examination officials
explained that a few large audits produced the peaks in 1991 and
1993.  In 1994 constant dollars, however, recommended taxes decreased
4 percent between 1988 and 1994.  (Refer to table II.5 for current
dollar data on additional taxes recommended and table V.1 for 1994
constant dollars.)

6.  Additional recommended taxes per return:  As with recommended
taxes, this ratio increased overall and for both categories of
corporations.  It rose 12 percent from about $160,000 in 1988 to
$180,000 in 1994 (after reaching $242,000 in 1993).  Higher asset
corporations drove this increase as its ratio increased 38 percent
after declining in 1992 and 1994.  Although IRS audited fewer of
their returns, IRS recommended a relatively higher amount of taxes. 
In 1994 constant dollars, recommended taxes per return decreased 7
percent between 1988 and 1994.  (Refer to table II.6 and table V.3
for details.)

7.  Additional recommended taxes per audit hour:  The overall ratio
decreased 7 percent from $1,409 in 1988 to $1,313 per hour in 1994. 
Although the ratio rose for corporations with higher assets, this
rise was more than offset by a declining ratio for those with lower
assets.  In 1994 constant dollars, the overall ratio decreased 23
percent from 1988 to 1994.  (Refer to tables II.7 and V.4.)

IRS Examination officials offered reasons for the increase in direct
audit hours outpacing increases in audit coverage and recommended
taxes after 1988.  IRS has been auditing more complex returns and
using more IRS specialists.  Both steps took more time and reduced
coverage.  Also, many auditors have needed more time as they
gradually shifted from auditing corporate tax shelters to auditing
the whole large corporation return.  On the other hand, recommended
amounts can be reduced by economic downturns.  They also cited 1986
tax law changes that took away audit issues (e.g., investment tax
credit) with relatively high yield for a small time investment and
that lowered corporate tax rates, affecting additional recommended
taxes in later years. 

We also analyzed trends in audit closures.  The large corporations
agreed with higher portions of recommended tax amounts from 1988 to
1994 (34 percent by 1994), but they continued to appeal most amounts
(66 percent by 1994).  Although they had similar trends over the 7
years, higher asset corporations agreed, on average, with 21 percent
of all recommended amounts while lower asset corporations agreed with
33 percent (see table II.9).  Even so, large corporations
increasingly agreed with most audits (52 percent for 1994) that
recommended taxes (see table II.8).  In sum, they tended to agree
with small tax amounts recommended in many audits but appeal larger
amounts recommended in fewer audits. 

Audits that end with no change to taxes owed could have adjustments
(e.g., reduced a reported net operating loss but not enough to
produce a tax liability) or no adjustments.  IRS views the former as
productive and the latter as unproductive.  The no-change trends
differed.  Those with adjustments dropped from 28 percent to 16
percent; those without adjustments increased from 8 to 16 percent. 
The without adjustments rate had reached 18 percent for lower asset
corporations and 10 percent for those with higher assets by 1994. 

IRS Examination officials said they would like to see the no-change
rate without adjustments fall below 10 percent.  They believed that
this rate will start falling as IRS closes ongoing audits that they
viewed as more productive.  For example, they believed that their
investment in audits of more complex returns will start shifting more
no-change audits to audits that recommend taxes.  Also, more auditors
have now learned how to audit large corporations, not just tax
shelters.  Even so, these officials still want better systems for
selecting and classifying (i.e., finding issues that need to be
audited) returns.  In recognizing this need, they convened task
forces during 1994 to overcome such problems with large corporation
audits.  These task forces are slated to last through 1996. 

See appendix II for detailed information about these trends and IRS'
explanations and appendix V for trends for recommended tax amounts in
constant dollars. 


   ASSESSMENT RATE
------------------------------------------------------------ Letter :5

In tracking taxes recommended for large corporations from 1988
through 1994, we found that the final assessment had been recorded
for $8.6 billion of $12 billion in net recommended taxes.  Our
computer match, involving about 56,000 audited returns, showed that
IRS assessed $2.3 billion of the $8.6 billion (27 percent) through
December 1994.  In computing this rate, we subtracted the tax refunds
recommended from the additional taxes recommended for these audited
returns. 

The assessment rate was similar for higher and lower asset
corporations--26 percent and 28 percent, respectively.  The rate,
however, differed widely by the four asset classes, ranging from 20
percent to 38 percent.  By IRS district, the assessment rate ranged
from over 100 percent to less than 1 percent. 

The reasons for these wide variations were not apparent in the IRS
databases we used to compute the assessment rates.  Our work has
shown that various factors can cause the rate to exceed 100 percent,
such as IRS Appeals assessing more taxes than recommended by the
auditors.  Also, the rates can drop whenever corporate claims for
refunds or net operating losses from other tax years reduce or offset
taxes that were recommended in the audit. 

Nor did IRS Examination officials know the reasons for the wide
variation in the rates.  They noted that the lowest rates occurred in
two regions and they are starting to pinpoint the reasons.  We also
plan to explore these reasons during a follow-on review. 

Our discussions with IRS Examination officials disclosed possible
reasons for low rates.  These officials pointed to nonaudit factors
that can lower the assessment rate, even if auditors supported the
taxes recommended.  They cited retroactive tax law changes, court
decisions that affect the recommended taxes, and other tax
abatements.  In addition, they said IRS Appeals can concede
recommended taxes to avoid the hazards of litigation or because the
corporation provides new information that swayed Appeals' decision;
this information could have dissuaded the auditors from recommending
the taxes.  These officials did not know the extent to which these
factors lowered the assessment rate, given limitations in IRS'
databases.  For this reason, our 1994 report on CEP recommended
corrections to IRS' databases. 

In sum, IRS Examination officials cautioned against misinterpreting
the assessment rates.  Because of these nonaudit factors, they
believed that the rates reflect more about the tax system and
economic fluctuations than the effectiveness of the audits.  Appendix
III provides details on assessment rates and IRS' explanations. 

We also computed the assessment rate for just the additional taxes
recommended (i.e., excluding audits recommending refunds).  That rate
equalled 38 percent.  IRS had estimated a similar assessment rate on
just the additional taxes recommended for audits closed in fiscal
years 1992 through 1994--36 percent. 

Regardless of which type of assessment rate is considered, we did not
attempt to track how much of the assessed taxes were ultimately
collected.  IRS Finance officials provided data indicating that IRS
collected 23 percent of the taxes recommended and 68 percent of the
taxes assessed as of July 1995 for the audits closed in fiscal years
1992 through 1994.  IRS based these results on data being tracked in
a new system.  We plan to analyze the data and methodology being used
in this system during the second phase of our work. 

Whenever audits recommend additional taxes that go unassessed, IRS
can miss opportunities to invest audit resources more productively,
and large corporations can incur more costs to challenge those
recommendations.  Data on many of these costs were not available. 
Using only the direct audit costs, we calculated that IRS recommended
$56 and assessed $15 in taxes for each dollar directly spent on
auditing large corporations from 1988 through 1994.\4

These calculations exclude indirect audit costs (e.g., overhead), IRS
costs outside of audits (e.g., appeals and litigation processes to
settle on assessed tax amounts), and corporations' costs.\5

It is important to recognize that these ratios provide just one
indicator of IRS' audit activities.  The ratios do not account for
the costs and taxes associated with what IRS calls revenue
protection.  For example, IRS may audit various corporate claims for
tax refunds to determine whether the claims are proper.  In doing so,
IRS protects the government's revenue.  Auditors disallowed $202
million in claims by large corporations during 1994 in addition to
the $1.9 billion they recommended in taxes.  In 1991--the first year
for which IRS tracked protected tax revenue--IRS auditors denied $212
million of these claims. 

To provide perspective, we computed a similar ratio for all IRS
audits.  Although not readily available for assessed taxes, data were
available in IRS' 1996 budget to compute the ratio of recommended
taxes to the costs of all IRS audits.  Our computations showed that
the ratio has been about $16 in recommended taxes to $1 in audit
costs (including indirect costs) for recent years. 


--------------------
\4 We calculated these ratios using the $153 million in direct audit
costs alone compared to recommended amounts of $8.6 billion and
assessed amounts of $2.3 billion. 

\5 In addition to nonaudit costs, the calculations excluded interest
that IRS assessed; assessed interest amounts roughly equaled assessed
tax amounts.  Further, the assessed tax ratio excluded the direct
audit costs associated with 28 percent of all recommended taxes over
the 7 years--about $3.4 billion--for which the assessed amounts have
not been finalized. 


   PROFILES OF LARGE CORPORATIONS
------------------------------------------------------------ Letter :6

According to 1992 income tax returns, over 60 percent of the large
corporations were engaged in manufacturing or in the
finance/insurance industry.  This profile was similar for both the
audited and nonaudited large corporations.  Audited large
corporations, however, tended to report higher amounts, on average,
of total income, taxable income, and income tax liability. 

Whether audited or not, large corporations tended to claim the
possessions tax credit more frequently than other tax credits;\6 57
percent of $6.4 billion in tax credits claimed was for the
possessions tax credit.  Appendix IV provides more details on the
profile of large corporations for 1992. 


--------------------
\6 Section 936 of the tax code provides a tax credit that equals the
full amount of the U.S.  income tax liability on income earned by
U.S.  firms from operations in U.S.  possessions.  This credit is
referred to as the possessions tax credit. 


   IRS COMMENTS AND OUR EVALUATION
------------------------------------------------------------ Letter :7

We requested comments on a draft of this report from the IRS
Commissioner, and we received comments from her representatives at a
meeting on August 9, 1995.  These IRS officials included the
Assistant Commissioner for Examination and his staff that oversee
audits of large corporations as well as staff from IRS' Office of
Legislative Affairs.  While generally agreeing with the trends we
analyzed, these officials had comments on our draft.  In addition to
technical comments that we have incorporated where appropriate, they
offered comments on three major issues. 

First, they pointed to various efforts undertaken to correct problems
with large corporation audits.  The major effort entails studying
ways to improve the selection of returns for audit.  Our letter now
refers to these efforts. 

Second, they suggested explanations for some trends.  For example,
they offered various reasons for the increases in audit coverage and
additional recommended taxes lagging behind the increase in direct
audit time.  They noted that IRS has been investing time in auditing
more complex issues and in using IRS specialists.  They viewed this
investment as necessary and as likely to pay off soon.  They also
cited tax law changes in 1986 and the transition in the early 1990s
from auditing corporate tax shelters to all large corporate tax
issues.  Both factors had dampening effects on recommended tax
amounts after 1988, according to these officials.  They suggested
that these factors, in combination with auditing more complex
returns, also contributed to IRS closing more audits with neither
changes to taxes owed nor adjustments to taxable income.  While we
did not validate IRS' suggested explanations, we have added them to
the letter and related appendixes. 

Third, they asked for clarification on the 27 percent assessment
rate.  Although our draft report had not labelled this rate as a
measure of audit effectiveness, they wanted cautions noted.  They
said the rate should not be used as such a measure because of
nonaudit factors (e.g., net operating losses from other tax years
that offset audit yield).  They did not know the extent to which
these factors affected the rate but they believed that the rate was
just as likely to be the product of the tax and economic systems,
which they have little control over, rather than of the audits.  We
have added their comments about the potential effects of these
nonaudit factors on the rates. 


---------------------------------------------------------- Letter :7.1

We are sending copies of this report to the Senate Committee on
Finance, the House Committee on Ways and Means, and other interested
parties.  Major contributors to this report are listed in appendix
VI.  If you or your staff have any questions concerning this report,
please contact me at (202) 512-5407. 

Sincerely yours,

Jennie S.  Stathis
Director, Tax Policy
 and Administration Issues


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

Our objectives were to (1) analyze audit trends for large
corporations not in the Coordinated Examination Program (CEP) for
fiscal years 1988 through 1994, (2) compute the portion of taxes
recommended in audits that were actually assessed, and (3) develop
and compare profiles of audited large corporations with those not
audited. 

The Internal Revenue Service (IRS) defines large corporations as
those reporting assets of $10 million or more on their income tax
returns.  IRS divides large corporations into four asset classes as
follows: 

(1)assets of $10 million to less than $50 million,
(2)assets of $50 million to less than $100 million,
(3)assets of $100 million to less than $250 million, and
(4)assets of $250 million and over. 

For these corporations, our analyses focused on data from Forms 1120
(U.S.  Corporation Income Tax Return) and other related corporate
returns, except for nontaxable returns such as the Form 1120-S. 
These related income tax returns included the following: 

(1)1120-L (U.S.  Life Insurance Company Income Tax Return),
(2)1120-PC (U.S.  Property and Casualty Insurance Company Income Tax
Return),
(3)1120 Consolidated income tax return,
(4)1120L Section 594/1504c income tax return for U.S.  life insurance
companies,
(5)1120-PC Section 1504c income tax return for U.S.  property and
casualty insurance companies,
(6)1120 Section 594/1504c income tax return for U.S.  corporations. 

Our analyses of audit trends, assessment rates, and the profiles
excluded large corporations in CEP.  We excluded CEP corporations
from the profile information using IRS data on CEP. 

We asked IRS Examination officials at the National Office to review
our analysis of the audit trends and assessment rates and to provide
any explanations.  We have summarized their comments throughout this
report. 


   ANALYZING AUDIT TRENDS
--------------------------------------------------------- Appendix I:1

To analyze audit trends, we used IRS' Audit Information Management
System (AIMS) data.  This database includes records from all audits
closed during a given fiscal year.  We reconciled totals from this
database to totals in IRS' annual report.\1

For audits closed from fiscal years 1988 through 1994, we obtained
AIMS data on additional tax recommended, tax decreases recommended,
returns audited, and direct audit hours spent on returns.  We then
calculated such measures as tax recommended per return, tax
recommended per hour, and audit hours per return.\2 Appendix II
reports these trends, using current dollars for the recommended tax
amounts.  Appendix V reports those trends in constant dollars. 

To calculate audit coverage, we used IRS' method of dividing the
number of audits completed in a given fiscal year by the number of
returns filed the previous year.  We also computed IRS' direct costs
for auditing these returns.  For audited corporations by asset class,
we applied the average cost IRS calculated for fiscal years 1991 and
1992 for each staff year that IRS directly spent on these audits.  We
adjusted the costs to current dollars for the specific fiscal year of
the audit to obtain the average cost for each of the 7 years we
analyzed. 

We also analyzed the ways in which IRS closed audits of the tax
returns.  If IRS recommended additional taxes, large corporations
could agree to pay or appeal these taxes.\3 If IRS did not recommend
such taxes, we analyzed how often IRS closed these no-change audits
without any audit adjustments or with adjustments.\4


--------------------
\1 AIMS data are the most reliable data available on audit results. 
Other than our reconciliation effort, we did not assess the
reliability of the AIMS data. 

\2 The term "recommended tax" refers to the amount of additional tax
that an auditor concludes should have been paid together with any
associated penalties.  Our analysis focused on recommended taxes. 
Penalties involved small amounts (less than 10 percent) of the
recommended taxes.  Interest amounts were not reflected in IRS' data
on audit results. 

\3 In analyzing closures, we treated "defaults" as part of "agreed"
closures.  Defaults arise when the corporation neither formally
agrees or appeals, meaning the recommended tax becomes assessed. 
Over the 7 years, defaults tended to be a small percentage of total
additional taxes recommended, ranging from 0 to 9 percent. 

\4 These adjustments can affect the amount of income, deductions, and
credits on tax returns but be offset by operating losses or excess
credits.  These adjustments can result in additional taxes in the
future because of carryover provisions in tax laws. 


   COMPUTING THE ASSESSMENT RATE
--------------------------------------------------------- Appendix I:2

To compute the assessment rate--the percentage of recommended taxes
ultimately assessed after audits--we did a computer data match of
corporate income tax returns between two IRS databases.  For all
closed audits in our populations, we matched the recommended tax
assessments recorded on AIMS to the actual tax assessments recorded
on the Business Master File (BMF).  In addition to the assessed tax
liabilities, the BMF contains information on taxable income, taxes
not yet paid, penalties, interest, payments, refunds, and audit
actions for business tax returns.  In both systems, each record
contains the taxpayer identification number (TIN), tax year, and
return type.\5

To use BMF data, we eliminated all BMF records of tax returns that
had no audit adjustment code.  We also eliminated all records with
audit transactions that were posted before fiscal year 1988.  Because
our AIMS data covered fiscal years 1988 through 1994, none of these
audit adjustments could have been posted on BMF before fiscal year
1988. 

Also, we applied our criterion of a "completed audit." We defined
this term as the period in which IRS made at least one tax adjustment
resulting from an audit, followed by an audit release indicator.  As
the starting point, we used the last day of the previous audit period
or, if not present, the date that IRS posted the return.  The BMF
audit release indicator identified the end of an audit.  We added 30
calendar days to the audit release date to identify late posting
audit adjustments.  IRS also does this adjustment on its new
Enforcement Revenue Information System to match tax adjustments to
taxes recommended. 

Using the 25,395 taxpayers identified in AIMS data for fiscal years
1988 through 1994, we were able to match 22,679 TINs to BMF.  For
these TINs, we obtained records for 56,146 returns for various tax
years ranging from 1964 to 1993.  AIMS has the recommended tax
adjustments for each closed audit.  We dropped records that showed
recommended taxes of $1 because some IRS districts use this amount
if, for some reason, they must close the case on AIMS for a second
time. 

Across the AIMS and BMF data, we sought the same corporate TINs and
same audited tax years for audits closed during fiscal years 1988
through 1994.  We matched the AIMS data on recommended assessments
from these audits to BMF data on actual assessments for these audits
up through December 1994.  We then analyzed the assessment rate by
variables such as the asset size of the corporation and IRS district
office that did the audit. 


--------------------
\5 We did not validate data recorded in BMF with source documents. 


   PROFILING LARGE CORPORATIONS
--------------------------------------------------------- Appendix I:3

To develop a profile of the large corporations, we obtained the 1992
Statistics of Income (SOI) file for corporations--the most recent
file when we did our work.  We eliminated CEP corporations as we did
for our other analyses.  We selected the large corporations by using
our criteria for return type and the asset size.  We matched these
data with AIMS data to divide the large corporation population into
audited and nonaudited groups.  Table I.1 shows the SOI universe and
populations for each of these steps. 



                               Table I.1
                
                 SOI Universe of Large Corporations for
                                  1992

Groups of SOI corporate returns                      Number of returns
--------------------------------------------------  ------------------
Total corporate returns                                         80,822
Returns left after excluding CEP corporations                   79,108
Returns left after applying return type and asset             24,604\a
 size criteria
Returns that were not audited                                 15,666\b
Returns that were audited                                      8,938\c
----------------------------------------------------------------------
\a This number is our universe of large corporate returns.  The
weighted value equals 39,155 returns. 

\b This number is our universe of nonaudited large corporate returns. 
The weighted value equals 25,600 returns. 

\c This number is our universe of audited large corporate returns. 
The weighted value equals 13,555 returns. 

Sampling errors associated with our SOI estimates are less than 5
percent at the 95 percent confidence level, except for the following
items: 

  For audited lower asset size corporations claiming the net
     operating loss deduction, the sampling errors were $1.3 billion
     + 6.8 percent for the net operating loss deduction claimed and
     $1 million + 5.9 percent for the average deduction claimed. 

  For the nonaudited corporations, the average Foreign Tax Credit
     claimed as shown in table IV.2 had a sampling error of $1.593
     million + 5.8 percent. 

  For the other tax credits reported in table IV.2, the sampling
     errors exceeded 5 percent for both audited and nonaudited
     corporations.  The sampling errors for the total amount claimed
     were $13.9 million + 5.7 percent and $11 million + 8.9 percent
     for audited and nonaudited corporations, respectively.  The
     sampling error for the average amount claimed is as follows: 
     $179 thousand + 16.6 percent and $225 thousand + 21.4 percent
     for audited and nonaudited corporations, respectively. 


AUDIT TRENDS FOR LARGE
CORPORATIONS
========================================================== Appendix II

This appendix presents our analysis of IRS' audit results for large
corporations, using IRS' AIMS data.  We asked IRS Examination
officials to explain any major shifts in the trends.  The narrative
within this appendix reflects any explanations that these officials
provided. 

   Figure II.1:  Number of Large
   Corporation Returns Audited,
   Fiscal Years 1988 Through 1994

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS data. 



                                     Table II.1
                      
                        Number of Large Corporation Returns
                      Audited, Fiscal Years 1988 Through 1994



Fi
sc
al                $50 mil.               $100 mil.
ye    $10 mil.      < $100                  < $250   $250 mil.
ar  < $50 mil.        mil.    Subtotal        mil.      & over    Subtotal    Total
--  ----------  ----------  ==========  ----------  ----------  ==========  =======
19       4,659       1,893       6,552       1,823       1,687       3,510   10,062
 88
19       5,132       1,679       6,811       1,548       1,361       2,909    9,720
 89
19       6,210       1,712       7,922       1,528       1,184       2,712   10,634
 90
19       7,104       1,877       8,981       1,627       1,354       2,981   11,962
 91
19       6,399       1,849       8,248       1,536       1,287       2,823   11,071
 92
19       6,292       1,677       7,969       1,586       1,321       2,907   10,876
 93
19       5,829       1,615       7,444       1,612       1,336       2,948   10,392
 94
===================================================================================
Av       5,946       1,757       7,704       1,609       1,361       2,970   10,674
 e
 r
 a
 g
 e
-----------------------------------------------------------------------------------
Source:  GAO analysis of IRS data. 

From 1988 to 1994, the number of returns audited has increased
slightly, particularly from 1988 to 1991, when it peaked at 11,962
returns and then decreased to 10,392 by 1994.  IRS Examination
officials attributed the decline since 1991 to auditing more complex
returns and issues.  Doing so takes more time that could have been
spent on more returns.  Corporations with less than $50 million in
assets accounted for the overall increase in audits during the 7
years.  All other asset classes had decreases. 

Over the 7 years, the number of returns audited averaged 10,674.  Of
these audited returns, lower asset corporations filed 72 percent
(7,704 returns) on average. 

   Figure II.2:  Audit Coverage
   for Large Corporations, Fiscal
   Years 1988 Through 1994

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS data. 



                                     Table II.2
                      
                       Audit Coverage for Large Corporations,
                           Fiscal Years 1988 Through 1994

                                 (Percent coverage)



Fi
sc                               Audit                               Audit    Total
al                $50 mil.    coverage   $100 mil.                coverage    audit
ye    $10 mil.      < $100       lower      < $250   $250 mil.      higher  coverag
ar  < $50 mil.        mil.  asset size        mil.      & over  asset size        e
--  ----------  ----------  ----------  ----------  ----------  ----------  =======
19         16%         31%         18%         41%         51%         45%      23%
 88
19          17          29          19          38          44          40       22
 89
19          19          32          21          41          52          45       24
 90
19          24          33          25          39          50          43       28
 91
19          23          27          24          28          36          31       25
 92
19          23          24          23          28          33          30       25
 93
19          21          23          22          28          34          31       24
 94
Av         20%         29%         22%         36%         44%         39%      25%
 e
 r
 a
 g
 e
 \
 a
-----------------------------------------------------------------------------------
\a Averages are computed using actual data points and may not equal
the averages of the whole numbers in the columns. 

Source:  GAO analysis of IRS data. 

For 1988 through 1994, audit coverage changed as follows: 

  It increased from 18 percent to 22 percent, peaking in 1991, for
     lower asset corporations.  This increase stemmed from IRS doing
     more audits while the number of returns filed remained fairly
     constant over the 7 years.  In fact, the audit rate increased
     only for corporations with assets of $10 million to less than
     $50 million. 

  It decreased from 45 percent to 31 percent for the higher asset
     classes.  Their rate held fairly steady through 1991 but then
     dropped through 1994.  Over all 7 years, their coverage
     decreased because more returns were filed but fewer were
     audited.  IRS Examination officials said IRS has spent more time
     on complex audits since 1991. 

  It averaged 25 percent overall.  As asset size increased, so did
     the average coverage rate.  Over the four asset classes, the
     average rate ranged from 20 percent to 44 percent. 

   Figure II.3:  Direct Audit
   Hours for Large Corporations,
   Fiscal Years 1988 Through 1994

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS data. 



                                     Table II.3
                      
                            Direct Audit Hours for Large
                      Corporations, Fiscal Years 1988 Through
                                        1994

                                (Hours in thousands)



Fi
sc
al                $50 mil.               $100 mil.
ye    $10 mil.      < $100                  < $250   $250 mil.
ar  < $50 mil.        mil.  Subtotal\a        mil.      & over  Subtotal\a  Total\a
--  ----------  ----------  ==========  ----------  ----------  ==========  =======
19         328         159         487         241         414         655    1,142
 88
19         359         161         520         224         272         496    1,016
 89
19         462         165         627         213         226         440    1,067
 90
19         600         216         816         266         234         501    1,317
 91
19         569         218         787         236         308         545    1,331
 92
19         560         220         780         268         297         564    1,344
 93
19         569         224         793         284         345         629    1,422
 94
===================================================================================
Av         493         195         687         248         300         547    1,234
 e
 r
 a
 g
 e
-----------------------------------------------------------------------------------
\a Difference due to rounding. 

Source:  GAO analysis of IRS data. 

Time spent directly on the audit is measured in hours.\1 After
dropping from 1988 to 1989, the audit hours steadily increased about
40 percent from fiscal years 1989 through 1994.  This change
represents an increase from about 1 million to 1.4 million staff
hours.  IRS Examination officials attributed the increase to
investing in audits of more complex tax returns and issues. 

The increased hours primarily arose from doing more audits of lower
asset corporations over the 7 years (see fig.  II.1).  The audit time
for these corporations increased from 487,162 hours in 1988 to
793,115 hours in 1994 (63 percent) after peaking in 1991 and then
decreasing slightly through 1994.  Audit hours for higher asset
corporations decreased from 654,974 hours in 1988 to 628,851 hours in
1994 (4 percent); their hours decreased from 1988 to 1990 and then
increased steadily through 1994. 

   Figure II.4:  Direct Audit
   Hours per Return for Large
   Corporations, Fiscal Years 1988
   Through 1994

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS data. 



                                     Table II.4
                      
                      Direct Audit Hours per Return for Large
                      Corporations, Fiscal Years 1988 Through
                                        1994



Fi
sc                                                                            Audit
al                $50 mil.       Audit   $100 mil.                   Audit    hours
ye    $10 mil.      < $100   hours per      < $250   $250 mil.   hours per      per
ar  < $50 mil.        mil.      return        mil.      & over      return  return\
--  ----------  ----------  ----------  ----------  ----------  ----------  -------
19          70          84          74         132         246         187      114
 88
19          70          96          76         145         200         171      105
 89
19          74          96          79         140         191         162      100
 90
19          84         115          91         164         173         168      110
 91
19          89         118          95         154         239         193      120
 92
19          89         131          98         169         224         194      124
 93
19          98         139         107         176         258         213      137
 94
Av          83         111          89         154         220         184      116
 e
 r
 a
 g
 e
 \
 a
-----------------------------------------------------------------------------------
\a Averages are computed using actual data points and may not equal
the averages of the whole numbers in the columns. 

Source:  GAO analysis of IRS data. 

On average, IRS auditors spent twice as long auditing returns from
higher asset corporations compared to those with lower assets--184
hours versus 89 hours per return, respectively. 

From 1988 to 1994, the direct audit hours per return increased 43
percent (74 to 107 hours) for lower asset corporations, 14 percent
(187 to 213 hours) for higher asset corporations, and 21 percent (114
to 137 hours) overall.  These increases match the increases in direct
audit hours (see table II.3) and in audited returns for lower asset
corporations and all large corporations (see table II.1). 

However, the number of direct audit hours and audited returns
decreased for higher asset large corporations.  The increase in
direct audit hours per return results from a larger decrease in the
number of audited returns (see table II.1) compared to the number of
audit hours (see table II.3). 

IRS Examination officials said they expected the upward trend in
audit hours per return to continue as IRS does more complex audits. 
They also cited other reasons.  The lack of training and experience
in auditing an entire large corporation return instead of just
corporate tax shelters has added time. 

   Figure II.5:  Additional Taxes
   Recommended for Large
   Corporations, Fiscal Years 1988
   Through 1994

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS data. 



                                     Table II.5
                      
                       Additional Taxes Recommended for Large
                      Corporations, Fiscal Years 1988 Through
                                        1994

                               (Dollars in millions)



Fi
sc
al                $50 mil.               $100 mil.
ye    $10 mil.      < $100                  < $250   $250 mil.
ar  < $50 mil.        mil.  Subtotal\a        mil.      & over  Subtotal\a  Total\a
--  ----------  ----------  ==========  ----------  ----------  ==========  =======
19        $402        $182        $584        $303        $723      $1,026   $1,610
 88
19         317         194         511        $246         686         932    1,444
 89
19         356         241         598        $567         792       1,359    1,957
 90
19         610         223         832        $449       1,074       1,523    2,356
 91
19         594         298         892        $498         866       1,364    2,256
 92
19         543         248         791        $506       1,331       1,837    2,629
 93
19         481         200         680        $429         758       1,186    1,867
 94
===================================================================================
Av        $472        $227        $698        $428        $890      $1,318   $2,017
 e
 r
 a
 g
 e
-----------------------------------------------------------------------------------
\a Difference due to rounding. 

Source:  GAO analysis of IRS data. 

For all large corporations, additional recommended taxes grew 16
percent from about $1.6 billion in 1988 to about $1.9 billion in
1994; except for 1989, these amounts had increased through 1993 and
then dropped in 1994.  IRS Examination officials did not know the
reasons for the 1994 decrease.  Both higher and lower asset
corporations had similar increases over the 7 years but those with
higher assets always accounted for the bulk of the additional tax
amounts. 

For higher asset corporations, recommended taxes peaked in
1993--about double the amount from 1989--but then decreased 35
percent by 1994.  IRS Examination officials attributed the big
increases in 1991 and in 1993 to a few large dollar audits.  For
lower asset corporations, recommended taxes increased 53 percent from
1988 to 1992 but then decreased 24 percent through 1994. 

All four asset classes had percentage increases from 1988 to 1994 in
taxes recommended; the $100 million to less than $250 million class
had the greatest increase (42 percent), while the $250 million and
over class had the smallest (5 percent).  Each class also had
fluctuations over the 7 years and different peak years, ranging from
1990 to 1993. 



                                    Table II.5.1
                      
                        Net Additional Taxes Recommended for
                       Large Corporations, Fiscal Years 1988
                                    Through 1994

                               (Dollars in millions)



Fi
sc
al                $50 mil.               $100 mil.
ye    $10 mil.      < $100                  < $250   $250 mil.
ar  < $50 mil.        mil.  Subtotal\a        mil.      & over  Subtotal\a  Total\a
--  ----------  ----------  ==========  ----------  ----------  ==========  =======
19        $354        $137        $491        $266        $571       $ 837   $1,328
 88
19         245         155         399         191         583         774    1,173
 89
19         236         184         420         521         731       1,252    1,672
 90
19         520         169         689         367         977       1,343    2,033
 91
19         520         259         778         412         770       1,182    1,960
 92
19         421         211         632         440       1,159       1,598    2,231
 93
19         423         150         572         372         618         990    1,562
 94
===================================================================================
Av        $388        $181        $569        $367        $773      $1,149   $1,708
 e
 r
 a
 g
 e
-----------------------------------------------------------------------------------
\a Difference due to rounding. 

Source:  GAO analysis of IRS data. 

IRS' audit mission is to determine the correct tax liability.  This
includes determining additional taxes that taxpayers owe or that
should be refunded to the taxpayers. 

Although IRS collects the data, IRS reports on audit results did not
offset recommended taxes by recommended tax refunds.  Our analysis of
IRS data showed that subtracting refunds from reported additional
taxes recommended would reduce additional taxes between 8 to 34
percent over the 7 years. 

   Figure II.6:  Additional Taxes
   Recommended per Return for
   Large Corporations, Fiscal
   Years 1988 Through 1994

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS data. 



                                     Table II.6
                      
                      Additional Taxes Recommended per Return
                        for Large Corporations, Fiscal Years
                                 1988 Through 1994

                               (Dollars in thousands)



Fi
sc                             Taxes       $100                  Taxes       Taxes
al   $10 mil.   $50 mil.  recommende       mil.       $250  recommende  recommende
ye      < $50     < $100       d per     < $250       mil.       d per       d per
ar       mil.       mil.      return       mil.     & over      return      return
--  ---------  ---------  ----------  ---------  ---------  ----------  ----------
19        $86        $96        $ 89       $166       $428        $292        $160
 88
19         62        116          75        159        504         320         149
 89
19         57        141          75        371        669         501         184
 90
19         86        119          93        276        793         511         197
 91
19         93        161         108        324        673         483         204
 92
19         86        148          99        319      1,008         632         242
 93
19         82        124          91        266        567         402         180
 94
==================================================================================
Av        $79       $141         $91       $266       $654        $444        $189
 e
 r
 a
 g
 e
 \
 a
----------------------------------------------------------------------------------
\a Averages are completed using actual data points and may not equal
the averages of the whole numbers in the columns. 

Source:  GAO analysis of IRS data. 

Over the 7 years, the additional taxes recommended per audited return
averaged about $189,000 for all large corporations.  Audits of higher
asset corporations drove this average; these audits averaged about
$444,000. 

A comparison of 1988 to 1994 showed that the ratio of recommended
taxes per return increased for lower and higher asset corporations
but at different rates and with different fluctuations, as follows: 

  For lower asset corporations, the ratio increased just 3 percent. 
     This ratio increased about 43 percent between 1990 and 1992 and
     then decreased through 1994. 

  For higher asset corporations, the ratio increased 38 percent from
     1988 to 1994 after fluctuations.  This ratio increased from 1988
     to 1990, flattened out for 1990 through 1992, increased in 1993,
     and then dropped 36 percent in 1994. 

As noted after table II.5, a few large cases drove the 1991 and 1993
results according to the IRS officials. 

   Figure II.7:  Additional Taxes
   Recommended per Direct Audit
   Hour for Large Corporations,
   Fiscal Years 1988 Through 1994

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS data. 



                                     Table II.7
                      
                      Additional Taxes Recommended per Direct
                         Audit Hour for Large Corporations,
                           Fiscal Years 1988 Through 1994



Fi
sc                                         $100
al   $10 mil.   $50 mil.       Taxes       mil.       $250       Taxes       Taxes
ye      < $50     < $100  recommende     < $250       mil.  recommende  recommende
ar       mil.       mil.  d per hour       mil.     & over  d per hour  d per hour
--  ---------  ---------  ----------  ---------  ---------  ----------  ----------
19     $1,224     $1,147      $1,199     $1,259     $1,744      $1,566      $1,409
 88
19        882      1,208         983      1,096      2,525       1,879       1,420
 89
19        771      1,464         953      2,657      3,504       3,093       1,835
 90
19      1,016      1,032       1,020      1,685      4,581       3,041       1,789
 91
19      1,044      1,372       1,134      2,104      2,812       2,504       1,695
 92
19        969      1,131       1,014      1,892      4,488       3,256       1,956
 93
19        845        891         858      1,510      2,196       1,887       1,313
 94
==================================================================================
Av       $958     $1,165      $1,016     $1,730     $2,972      $2,410      $1,634
 e
 r
 a
 g
 e
 \
 a
----------------------------------------------------------------------------------
\a Averages are computed using actual data points and may not equal
the averages of the whole numbers in the columns. 

Source:  GAO analysis of IRS data. 

A comparison of 1988 to 1994 showed that taxes recommended per audit
hour decreased 7 percent.  This ratio increased from 1988 to 1990 but
then fluctuated through 1994.  It increased for higher asset
corporations and decreased for lower asset corporations.  More
specifically, this ratio: 

  increased for higher asset corporations because audits of those
     with assets of (1) $100 million to less than $250 million
     recommended more taxes for a proportionately smaller increase in
     audit hours, and (2) $250 million or more spent fewer audit
     hours to recommend a slight increase in the amount of taxes; and

  decreased for lower asset corporations because audits of those with
     assets of (1) $10 million to less than $50 million required more
     time to recommend less tax, and (2) $50 million to less than
     $100 million invested comparatively higher amounts of audit time
     to recommend higher tax amounts. 

IRS Examination officials pointed to a few large audits as major
contributors to the 1991 and 1993 results. 

   Figure II.8:  Percent of
   Returns Audited by Type of
   Closure for Large Corporations,
   Fiscal Years 1988 Through 1994

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS data. 



                               Table II.8
                
                 Percent of Returns Audited by Type of
                     Closure for Higher Asset Large
                Corporations, Fiscal Years 1988 Through
                                  1994


                                       No change   No change
                                            with     without
                                      adjustment  adjustment
Fiscal year       Appealed    Agreed           s           s   Total\a
----------------  --------  --------  ----------  ----------  ========
Total -lower asset size
----------------------------------------------------------------------
1988                   24%       39%         27%         10%      100%
1989                    20        44          24          11       100
1990                    20        46          21          13       100
1991                    19        46          19          16       100
1992                    18        49          16          17       100
1993                    17        50          15          18       100
1994                    16        51          16          18       100
Average                19%       47%         20%         15%      100%

Total -higher asset size
----------------------------------------------------------------------
1988                   33%       32%         29%          5%      100%
1989                    35        36          24           5       100
1990                    34        36          22           7       100
1991                    32        39          20           8       100
1992                    33        43          16           8       100
1993                    28        47          17           8       100
1994                    19        54          16          10       100
Average                31%       41%         21%          7%      100%

Total -all asset sizes
----------------------------------------------------------------------
1988                   27%       37%         28%          8%      100%
1989                    25        42          24           9       100
1990                    24        44          21          11       100
1991                    22        44          19          14       100
1992                    22        48          16          15       100
1993                    20        50          15          15       100
1994                    17        52          16          16       100
======================================================================
Average                22%       45%         20%         13%      100%
----------------------------------------------------------------------
\a Difference due to rounding. 

Source:  GAO analysis based on IRS data. 

Our analysis of how IRS closed audits of large corporations revealed
two distinct trends from 1988 through 1994.  These trends were
similar for lower and higher asset corporations. 

First, large corporations appealed a smaller percentage of the
returns that recommended additional taxes and agreed with a higher
percentage of these returns.  A comparison of 1988 to 1994 showed
that higher asset corporations reduced their appeal rate from 33
percent to 19 percent of the audited returns; this rate was 28
percent in 1993.  Lower asset corporations reduced this rate from 24
percent to 16 percent of the returns. 

Second, the two types of no-change rates moved in different
directions over the 7 years.  The rate with audit adjustments
steadily declined and the rate without audit adjustments slowly
increased.  The rate without adjustments increased for both types of
corporations over the 7 years--a 100-percent increase for higher
asset corporations and an 80-percent increase for lower asset
corporations.  Conversely, the rate with adjustments decreased by
over 40 percent for both higher and lower asset corporations. 

Further, a comparison of 1988 rates to 1994 rates showed that the
no-change rate without adjustments for lower asset corporations was
about twice the rate for higher asset corporations.  Over the 7
years, the overall no-change rate averaged 34 percent for lower asset
corporations and 28 percent for higher asset corporations. 

IRS Examination officials noted these increases in the no-change rate
without adjustments.  For all large corporations, this rate was 16
percent in 1994; this rate had been at or above 10 percent for lower
asset corporations and had reached 10 percent for higher asset
corporations by 1994.  These officials said a more satisfactory rate
would be less than 10 percent within all IRS regions.  They
recognized the need to be more selective in placing returns into the
audit stream.  They noted that this will only be accomplished by
universally using a process that better selects returns for audit and
that then identifies issues on those returns that need to be audited. 
They pointed to a task force that IRS convened in 1994 to develop the
universal process across IRS. 

   Figure II.9:  Percent of
   Additional Taxes Recommended by
   Type of Closure for Large
   Corporations, Fiscal Years 1988
   Through 1994

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS data. 



                               Table II.9
                
                Percent of Additional Taxes Recommended
                for Large Corporations by Asset Size and
                   Type of Closure, Fiscal Years 1988
                              Through 1994



                     Appeale          Appeale          Appeale
Fiscal year              d\a  Agreed      d\a  Agreed      d\a  Agreed
-------------------  -------  ------  -------  ------  -------  ------
1988                     70%     30%      76%     24%      74%     26%
1989                      62      38       75      25       70      30
1990                      76      24       88      12       85      15
1991                      61      39       87      13       78      22
1992                      70      30       82      18       77      23
1993                      64      36       76      24       72      28
1994                      64      36       67      33       66      34
======================================================================
Average                  67%     33%      79%     21%      75%     25%
----------------------------------------------------------------------
\a In 1991, IRS started to track partial agreements for appealed
returns.  The percent of appealed dollars includes these partial
assessments which is less than 6 percent of the total additional
taxes recommended. 

Source:  GAO analysis of IRS data. 

As with the trends in closing audited returns, a comparison of 1988
to 1994 showed that the large corporations appealed less and agreed
with more of the recommended tax amounts.  Unlike with the return
trends, the large corporations appealed the majority of these amounts
(about two-thirds by 1994). 

Further, the larger the asset size, the more likely that the large
corporation would appeal the recommended taxes rather than agree to
pay them.  On average, higher asset corporations appealed 79 percent
of the recommended tax amounts compared to 67 percent for those with
lower assets over the 7 years. 


--------------------
\1 A direct audit hour refers to time spent directly on the audit by
IRS technical staff.  It excludes training, leave, and time spent by
managers.  A direct audit staff year equals 2,000 hours. 


ASSESSMENT RATES FOR LARGE
CORPORATIONS
========================================================= Appendix III

Table III.1 shows our computation of the assessment rate on net tax
recommendations--recommended additional taxes less recommended tax
decreases.  IRS assessed $2.33 billion (27 percent) of the $8.64
billion in taxes recommended.\1 IRS made the assessments through
December 1994 for audits closed in fiscal years 1988 through 1994. 



                              Table III.1
                
                   Assessment Rate on Net Additional
                      Recommended Taxes for Large
                Corporations, Fiscal Years 1988 Through
                                  1994

                         (Dollars in millions)

                                         Taxes
Large corporation                   recommende       Taxes  Assessment
asset size                                   d    assessed        rate
----------------------------------  ----------  ----------  ----------
$10 mil.                                $2,089       $ 621      29.75%
 < $50 mil.
$50 mil.                                   888         220       24.78
 < $100 mil.
======================================================================
$10 mil.                                 2,977         841       28.27
 < $100 mil.
$100 mil.                                2,043         776       37.98
 < $250 mil
$250 mil. & over                         3,620         714       19.72
======================================================================
$100 mil. & over                         5,663       1,490       26.31
======================================================================
Total                                   $8,640      $2,331      26.98%
----------------------------------------------------------------------
Source:  GAO analysis of IRS data. 

The higher and lower asset corporations had similar assessment
rates--26 percent and 28 percent, respectively.  By asset class, the
rates ranged from 20 percent to 38 percent.  The 38 percent rate was
driven by audits in the Manhattan District Office.  Manhattan
accounted for $467.9 million (23 percent) of taxes recommended and
$372.7 million (48 percent) of taxes assessed in that asset class. 

IRS Examination officials itemized factors outside of the audits that
depressed the assessment rate.  For this reason, they cautioned
against using the rate to measure audit effectiveness.  They pointed
to net operating losses that large corporations carried over to
offset taxes recommended as well as offsets or reductions by claims
for refunds, abatements, and retroactive tax law changes and court
decisions.  They also said recommended taxes can be lost in Appeals
due to hazards of litigation and to large corporations withholding
tax data until then.  If IRS auditors had had these data, they would
have been less likely to recommend taxes.  These officials did not
know the extent to which these nonaudit factors affected the
assessment rate. 

IRS reports audit results by the gross recommended additional taxes. 
Table III.2 shows the assessment rate when recommended tax increases
and tax decreases are not netted. 



                              Table III.2
                
                  Assessment Rate on Gross Additional
                      Taxes Recommended for Large
                Corporations, Fiscal Years 1988 Through
                                  1994

                         (Dollars in millions)

                                         Taxes
Large corporation                   recommende       Taxes  Assessment
asset size                                   d    assessed        rate
----------------------------------  ----------  ----------  ----------
$10 mil.                                $2,548       $ 985      38.65%
 < $50 mil.
$50 mil.                                 1,164         464       39.87
 < $100 mil.
======================================================================
$10 mil.                                 3,713       1,449       39.03
 < $100 mil.\a
$100 mil.                                2,413       1,084       44.91
 < $250 mil
$250 mil. & over                         4,304       1,378       32.03
======================================================================
$100 mil. & over\a                       6,717       2,462       36.66
======================================================================
Total\a                                $10,430      $3,912      37.51%
----------------------------------------------------------------------
\a Difference due to rounding. 

Source:  GAO analysis of IRS data. 

Our analysis of just the gross additional taxes recommended showed a
higher assessment rate overall (38 percent) and by asset class (32
percent to 45 percent) compared to the net rate.  IRS has estimated
similar gross assessment rates using data from fiscal years 1992
through 1994 versus our 7-year period.  IRS' rate was 36 percent and
ranged from 24 percent to 54 percent by asset class. 

Table III.3 provides the net assessment rate for the districts that
recommended at least $100 million in additional taxes. 



                              Table III.3
                
                Districts With Over $100 Million in Net
                 Additional Taxes Recommended for Large
                  Corporations in Descending Order of
                   Assessment Rate, Fiscal Years 1988
                              Through 1994

                         (Dollars in millions)

                                Number
                                    of   Net taxes   Taxes
                                return  recommende  assess  Assessment
District Name                        s           d      ed        rate
------------------------------  ------  ----------  ------  ----------
1. Cleveland                     1,154        $117    $120     102.70%
2. Manhattan                     2,436         507     343       67.63
3. Greensboro, NC                  929         187     117       62.41
4. St. Paul                      1,162         175      78       44.92
5. Des Moines                      535         117      50       43.24
6. Atlanta                       1,123         228      96       42.20
7. Newark                        2,113         159      54       33.93
8. Detroit                       1,654         324     104       32.07
9. Milwaukee                       972         118      36       30.20
10. St. Louis                      801         163      47       28.55
11. Philadelphia                 1,363         106      25       24.03
12. Chicago                      1,631         139      31       22.01
13. Salt Lake City                 263         111      24       21.66
14. Dallas                       2,310         189      36       19.17
15. Houston                        796         204      32       15.54
16. Laguna Nigel, CA               948         336      39       11.73
17. Oklahoma City                  831         230      19        8.29
18. Los Angeles                    966         484      17        3.60
19. San Francisco                  549         611       6        0.90
20. Baltimore                    1,036         144       0        0.08
All other districts             21,843       1,815     584       32.18
Not allocable to a district     10,731       2,175     473       21.73
 office\a
======================================================================
Total\b                         56,146      $8,640  $2,331      26.98%
----------------------------------------------------------------------
\a IRS records did not identify a district office for all taxes
recommended.  The recommended taxes allocable to a district were
about $6.5 billion (75 percent). 

\b Difference due to rounding. 

Source:  GAO analysis of IRS data. 

Across IRS' 64 districts, 20 recommended net additional taxes of $100
million or more between fiscal years 1988 and 1994.  Three of the 20
districts--San Francisco, Manhattan, and Los Angeles--each
recommended about $500 million or more and together accounted for
$1.6 billion of the $8.6 billion recommended by all districts. 

As table III.3 shows, the assessment rates varied widely across the
districts.  The rates reached as high as about 103 percent to less
than 1 percent.  By including all but two IRS districts doing large
corporate audits, the lowest rate was negative 20 percent.\2

Assessment rates that exceed 100 percent indicate that appeals
officers assessed more taxes than recommended by revenue agents. 
This can occur when further adjustments increase tax liability while
the case is under Appeals' jurisdiction.  For example, liability
increases can occur when an adjustment on another tax year decreases
a net operating loss carryback deduction on the tax years being
appealed, a math error is found, or a taxpayer files an amended
return increasing tax liability. 

Negative assessment rates occur when the appeals officer not only
concedes all taxes recommended but also approves a tax refund because
the taxpayer filed a claim for refund or the appeals process reduced
the reported tax liability.  For example, the appeals officer can
decrease tax liability because of an error in the taxes recommended
or an increase in a loss carryback deduction from another tax year to
the tax year in Appeals. 

IRS Examination officials did not know the specific reasons for the
lowest rates.  They noted that two regions tended to account for
these lowest rates.  They planned to follow up with the regions to
uncover the reasons and see whether actions need to be taken. 

Using gross additional taxes recommended, the assessment rate and
ranking of the 20 districts changed slightly.  The rate ranges from
90 percent to 3 percent.  (See table III.4.)



                              Table III.4
                
                 Descending Order of Large Corporation
                 Assessment Rates for Gross Additional
                 Taxes Recommended for Those Districts
                With Over $100 Million in Net Additional
                  Taxes Recommended, Fiscal Years 1988
                              Through 1994

                         (Dollars in millions)

                                Number       Gross
                                    of       taxes   Taxes
                                return  recommende  assess  Assessment
District Name                        s           d      ed        rate
------------------------------  ------  ----------  ------  ----------
1. Cleveland                       985        $145    $130      89.88%
2. Manhattan                     2,059         723     537      74.32%
3. Greensboro, NC                  810         209     139      66.37%
4. Newark                        1,727         254     144      56.75%
5. St. Paul                      1,003         185      88      47.62%
6. Atlanta                         980         248     116      46.86%
7. Des Moines                      449         129      60      46.19%
8. Milwaukee                       799         138      61      44.41%
9. Philadelphia                  1,113         142      60      42.02%
10. Chicago                      1,316         214      87      40.84%
11. Dallas                       1,890         259     103      39.54%
12. Detroit                      1,475         358     135      37.59%
13. St. Louis                      661         182      65      35.53%
14. Salt Lake City                 215         119      33      27.68%
15. Houston                        668         228      53      23.33%
16. Baltimore                      871         179      27      15.36%
17. Oklahoma City                  690         246      33      13.52%
18. Laguna Nigel, CA               839         365      48      13.16%
19. Los Angeles                    806         524      53      10.11%
20. San Francisco                  460         628      20       3.26%
All other districts             18,234       2,347   1,026      43.73%
Not allocable to a district
 office\a                        8,543       2,609     893      34.22%
======================================================================
Total\b                         46,593     $10,430  $3,912      37.51%
----------------------------------------------------------------------
\a IRS records did not identify a district office for all taxes
recommended.  The gross recommended taxes allocable to a district
were about $7.8 billion (75 percent). 

\b Difference due to rounding. 

Source:  GAO analysis of IRS data. 


--------------------
\1 This rate approximates our computed CEP rate--22 percent ($7.1
billion of $32.4 billion).  Both rates exclude interest.  For CEP,
interest amounted to $3.3 billion; for other large corporations,
interest amounted to about $2.2 billion--$1.6 billion for those with
higher assets and $0.6 billion for those with lower assets. 

\2 We excluded two districts whose assessment rates exceeded negative
100 percent because all of the net recommended dollars were refunds
and IRS eventually refunded even more; the IRS databases did not
indicate the reasons.  Regardless, the recommended refunds totaled
less than 0.1 percent of all net taxes recommended. 


PROFILE OF LARGE CORPORATIONS
========================================================== Appendix IV

To develop a profile of large corporations, we used SOI data on large
corporations that filed income tax returns for 1992.  We split our
profile into large corporations that were audited and not audited,
using AIMS data.  Within that framework, the elements we profiled
included the type of industry, asset size, reported income and tax,
and types of tax credits claimed. 

   Figure IV.1:  Audited Large
   Corporations by Type of
   Industry, 1992

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS' SOI data. 

   Figure IV.2:  Nonaudited Large
   Corporations by Type of
   Industry, 1992

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS' SOI data. 

The majority of audited large corporations were engaged in finance,
insurance and real estate (34 percent); manufacturing (28 percent);
and wholesale trade (13 percent). 

The industry profile differs for the large corporations that were not
audited.  A higher percentage of them were involved in
finance/insurance and real estate (52 percent), and a lower
percentage were in manufacturing (17 percent) and wholesale trade (8
percent).  The third-ranking industry involved services (9 percent). 

For both audited and nonaudited returns of the corporations involved
in finance, insurance and real estate, the majority involved banks
and credit agencies. 

In the manufacturing industry for audited returns, the corporations
primarily manufactured electronic equipment, fabricated metals (such
as metal cans and shipping containers), food, and chemicals.  For
returns not audited, the manufacturing corporations had a similar
industry profile to those that were audited.  They were largely
involved in the same industries. 

   Figure IV.3:  Percent of Large
   Corporations by Asset Size,
   1992

   (See figure in printed
   edition.)

Source:  GAO analysis of IRS' SOI data. 

In 1992, 79 percent of the large corporations reported assets of $10
million to $100 million (lower asset size), and 21 percent reported
$100 million or more (higher asset size).  Over 60 percent reported
less than $50 million in assets. 

For all corporations, the average asset size was $131 million.  The
average asset size was $32 million for lower asset corporations and
$510 million for higher asset corporations. 

Comparing those audited versus not audited, the results were similar. 
For example, 75 percent of the audited corporations and 81 percent of
the nonaudited corporations were in the lower asset group; the rest
were higher asset corporations.  On average, the audited corporations
reported $136 million in assets while those not audited reported $128
million in assets. 



                               Table IV.1
                
                 Income and Tax Reported by Audited and
                 Nonaudited Large Corporations by Asset
                               Size, 1992

                         (Dollars in thousands)

                                       Average                 Average
                           Amounts     amounts     Amounts     amounts
----------------------  ----------  ----------  ----------  ----------
Higher asset size -$100 million & over
----------------------------------------------------------------------
                         Audited\a   Audited\a         Not         Not
                                                 audited\b   audited\b
Total income            $308,701,2     $92,787  $412,707,1     $86,904
                                28                      83
Taxable income          36,969,665      11,112  39,261,772       8,267
Income tax              12,576,614       3,780  13,291,383       2,799
Net tax                 11,173,474       3,358  10,617,493       2,236

Lower asset size -$10 million to $100 million
----------------------------------------------------------------------
                         Audited\c   Audited\c         Not         Not
                                                 audited\d   audited\d
Total income            $143,372,3     $14,018  $184,122,0      $8,830
                                54                      17
Taxable income          15,038,065       1,470  15,504,506         744
Income tax               5,105,788         499   5,233,269         251
Net tax                  4,868,725         476   4,773,347         229
----------------------------------------------------------------------
\a Represents 3,327 corporations. 

\b Represents 4,749 corporations. 

\c Represents 10,228 corporations. 

\d Represents 20,851 corporations. 

Source:  GAO analysis of IRS' SOI data. 

In dollar amounts, the nonaudited large corporations reported more
total income, taxable income, and income tax than the audited
corporations.  However, the audited large corporations reported
higher average amounts in these categories.  These higher average
amounts varied by asset group.  For example: 

  Higher asset corporations that were audited reported higher average
     amounts in these categories than those not audited. 

  Lower asset corporations that were audited reported much higher
     average amounts in these categories than those not audited.  The
     reported average amounts by the audited group usually doubled or
     almost doubled these amounts for the nonaudited group, except
     for total income.  For total income, the audited group reported
     about 59 percent more on average. 

Further analysis uncovered other results for 1992, as follows. 

  Among all large corporations, the approximate $31.5 billion in
     reported net tax was about 30 percent of the approximate $106.8
     billion in reported taxable income. 

  The percentage of returns reporting zero taxable income and zero
     tax was lower for audited returns compared to nonaudited
     returns.  For example, 22 percent of audited returns reported
     zero net tax compared to 36 percent for nonaudited returns.\1

We also analyzed the net operating loss deduction that large
corporations claimed to reduce their 1992 taxable income.  Audited
higher asset corporations claimed about $2.5 billion (average of
about $4.9 million), and those not audited claimed about $6.6 billion
(average of about $7.8 million).  Among lower asset corporations,
those audited claimed about $1.3 billion (average of about $1
million), and those not audited claimed about $3.9 billion (average
of about $0.9 million). 



                               Table IV.2
                
                   Tax Credits Claimed by Audited and
                  Nonaudited Large Corporations, 1992

                         (Dollars in thousands)

                            Number       Total     Percent     Average
                          claiming      amount    of total      amount
Type of credit              credit     claimed      amount     claimed
----------------------  ----------  ----------  ----------  ----------
Audited corporations\a
----------------------------------------------------------------------
Possessions tax credit          61  $1,037,698         47%     $17,011
Foreign tax credit             995     653,050         30%         656
General business             1,577     247,092         11%         157
 credit
Alternative minimum          1,412     236,029         11%         167
 tax credit
Other\b                         78      13,968          1%         179
Total credit                 3,359  $2,187,837        100%       $ 651

Nonaudited corporations\c
----------------------------------------------------------------------
Possessions tax credit         174  $2,628,703         62%     $15,107
Foreign tax credit             697   1,110,210          26       1,593
General business             1,660     232,950           5        $140
 credit
Alternative minimum          1,980     257,573           6        $130
 tax credit
Other\b                         49      11,009          \d        $225
======================================================================
Total credit                 3,875  $4,240,444        100%      $1,094
----------------------------------------------------------------------
\a Represents 13,555 corporations. 

\b Nonconventional fuel source credit and orphan drug credit. 

\c Represents 25,600 corporations. 

\d Less than 1 percent. 

Source:  GAO analysis of IRS' SOI data. 

For corporations audited, the possessions tax credit accounted for 47
percent of the total credits; the foreign tax credit represented
another 30 percent.  Both credits were claimed primarily by higher
asset corporations.  Among corporations not audited, the possessions
tax credit accounted for 62 percent, and the foreign tax credit
accounted for 26 percent of the total credits claimed. 

Because the possessions tax credit was claimed so much, we looked
more closely at which types of audited and nonaudited large
corporations claimed this credit.  The differences were minor, as
illustrated below. 

  Among audited corporations, manufacturers claimed 98 percent;
     manufacturers of chemicals/drugs and food claimed 79 percent of
     the total.  The majority of the corporations claiming this
     credit were in the higher asset group (90 percent). 

  Among nonaudited corporations, manufacturers claimed 96 percent;
     manufacturers of chemicals/drugs and instruments and related
     products claimed 80 percent of the total.  Further, 84 percent
     of the nonaudited corporations claiming the credit fell into the
     higher asset group. 


--------------------
\1 Of the 13,555 audited corporations, 2,956 reported zero net tax
(22 percent); of the 25,600 nonaudited corporations, 9,200 reported
zero net tax (36 percent). 


ADDITIONAL TAXES RECOMMENDED
PRESENTED IN CONSTANT DOLLARS
=========================================================== Appendix V



                                     Table V.1
                      
                       Additional Taxes Recommended Presented
                           in Constant Dollars for Large
                      Corporations, Fiscal Years 1988 Through
                                        1994

                               (Dollars in millions)



Fi
sc
al                $50 mil.               $100 mil.
ye    $10 mil.      < $100                  < $250   $250 mil.
ar  < $50 mil.        mil.  Subtotal\a        mil.      & over  Subtotal\a  Total\a
--  ----------  ----------  ==========  ----------  ----------  ==========  =======
19        $488        $221        $709        $368        $877      $1,245   $1,954
 88
19         368         226         594        $286         798       1,083    1,678
 89
19         397         269         665        $631         882       1,513    2,178
 90
19         654         239         892        $481       1,152       1,633    2,526
 91
19         620         311         931        $519         904       1,423    2,354
 92
19         554         254         808        $517       1,359       1,876    2,684
 93
19         481         200         680        $429         758       1,186    1,867
 94
===================================================================================
Av        $509        $246        $754        $462        $961      $1,423   $2,177
 e
 r
 a
 g
 e
-----------------------------------------------------------------------------------
Note:  Taxes recommended are presented in 1994 dollars. 

\a Difference due to rounding. 

Source:  GAO analysis of IRS data. 

In a comparison of 1988 to 1994, additional taxes recommended
decreased 4 percent overall and for lower asset corporations in 1994
constant dollars.  For higher asset corporations, the recommended
taxes decreased 5 percent from $1,245 million in 1988 to $1,186
million in 1994.  The greatest decrease of 14 percent occurred for
corporations with assets of $250 million and over ($877 million in
1988 to $758 million in 1994). 



                                     Table V.2
                      
                          Net Additional Taxes Recommended
                      Presented in Constant Dollars for Large
                      Corporations, Fiscal Years 1988 Through
                                        1994

                               (Dollars in millions)



Fi
sc
al                $50 mil.               $100 mil.
ye    $10 mil.      < $100                  < $250   $250 mil.
ar  < $50 mil.        mil.  Subtotal\a        mil.      & over  Subtotal\a  Total\a
--  ----------  ----------  ==========  ----------  ----------  ==========  =======
19        $430        $166        $596        $323        $692      $1,016   $1,611
 88
19         284         180         464         222         678         899    1,363
 89
19         262         205         467         580         814       1,394    1,861
 90
19         558         182         739         393       1,047       1,440    2,180
 91
19         542         270         812         430         803       1,233    2,045
 92
19         430         216         646         450       1,183       1,632    2,278
 93
19         423         150         572         372         618         990    1,562
 94
===================================================================================
Av        $418        $195        $614        $396        $834      $1,229   $1,843
 e
 r
 a
 g
 e
-----------------------------------------------------------------------------------
Note:  Taxes recommended are presented in 1994 dollars. 

\a Difference due to rounding. 

Source:  GAO analysis of IRS data. 

In 1994 constant dollars, net additional taxes recommended slightly
decreased 3 to 4 percent overall and for higher and lower asset
corporations in a comparison of 1988 to 1994 results.  Of the four
asset classes, only the audits of corporations with assets of $100
million to less than $250 million generated more net recommended
taxes--about 15 percent (from $323 million in 1988 to $372 million in
1994). 



                                              Table V.3
                               
                               Additional Taxes Recommended per Return,
                               Presented in Constant Dollars for Large
                               Corporations, Fiscal Years 1988 Through
                                                 1994

                                        (Dollars in thousands)



                                               Taxes                               Taxes       Taxes
                    $10 mil.    $50 mil.  recommende   $100 mil.              recommende  recommende
                           <           <       d per           <   $250 mil.       d per       d per
Fiscal year         $50 mil.   $100 mil.      return   $250 mil.    and over      return      return
----------------  ----------  ----------  ==========  ----------  ----------  ==========  ==========
1988                    $105        $117        $108        $202        $520        $355        $194
1989                      72         135          87         185         585         372         173
1990                      64         157          84         413         745         558         205
1991                      92         127          99         296         851         548         211
1992                      97         168         113         338         702         504         213
1993                      88         151         101         326       1,029         645         247
1994                      82         124          91         266         567         402         180
Average                  $86        $140         $98        $287        $706        $479        $204
----------------------------------------------------------------------------------------------------
Note:  Taxes recommended are presented in 1994 dollars. 

Source:  GAO analysis of IRS data. 

In 1994 constant dollars, a comparison of 1988 to 1994 showed that
the amount of additional taxes recommended per return has decreased
overall and for lower asset size corporations.  For higher asset size
corporations, this ratio increased 13 percent (about $355,000 in 1988
to $402,000 in 1994).  However, corporations with assets of $10
million to less than $50 million drove the overall change with a
decrease of 21 percent of recommended taxes per return (from about
$105,000 in 1988 to about $82,000 in 1994). 



                                              Table V.4
                               
                               Additional Taxes Recommended per Direct
                                  Audit Hour, Presented in Constant
                                Dollars for Large Corporations, Fiscal
                                       years 1988 Through 1994

                                        (Dollars in thousands)



                                               Taxes                               Taxes       Taxes
                    $10 mil.    $50 mil.  recommende   $100 mil.              recommende  recommende
                           <           <       d per           <   $250 mil.       d per       d per
Fiscal year         $50 mil.   $100 mil.  audit hour   $250 mil.    and over  audit hour  audit hour
----------------  ----------  ----------  ==========  ----------  ----------  ==========  ==========
1988                  $1,485      $1,392      $1,455      $1,528      $2,117      $1,901      $1,710
1989                   1,025       1,404       1,142       1,274       2,935       2,184       1,651
1990                     858       1,629       1,061       2,957       3,900       3,442       2,042
1991                   1,089       1,106       1,094       1,807       4,912       3,261       1,918
1992                   1,089       1,431       1,183       2,195       2,933       2,612       1,768
1993                     989       1,155       1,036       1,932       4,583       3,325       1,997
1994                     845         891         858       1,510       2,196       1,887       1,313
====================================================================================================
Average               $1,033      $1,262      $1,098      $1,865      $3,210      $2,601      $1,764
----------------------------------------------------------------------------------------------------
Note:  Taxes recommended are presented in 1994 dollars. 

Source:  GAO analysis of IRS data. 

A decrease occurred overall and for each asset class except for
corporations with assets of $250 million and over.  The overall
decrease from $1,710 in 1988 to $1,313 in 1994 was 23 percent. 
However, corporations with assets of $10 million to less than $50
million experienced the greatest decrease of 43 percent ($1,455 in
1988 to $858 in 1994). 


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

GENERAL GOVERNMENT DIVISION,
WASHINGTON, D.C. 

Thomas D.  Short, Assistant Director, Tax Policy and
 Administration Issues
Patricia H.  McGuire, Computer Specialist
Susan Baker, Computer Assistant

KANSAS CITY REGIONAL OFFICE

Cecelia M.  Ball, Project Manager
Royce L.  Baker, Issue Area Coordinator
Thomas N.  Bloom, Computer Specialist