Tax Administration: More Criteria Needed on IRS' Use of Financial Status
Audit Techniques (Letter Report, 12/30/97, GAO/GGD-98-38).

Pursuant to a congressional request, GAO reviewed the Internal Revenue
Service's (IRS) use of financial status audit techniques to: (1)
estimate how frequently IRS used financial status audit techniques in
audits closed in tax years prior to the 1994 initiative (1992 and 1993)
and in tax years following the 1994 initiative (1995 and 1996); (2)
consider how IRS' need to contact taxpayers for additional taxpayer
information when using financial status techniques might intrude on
taxpayers; (3) estimate the audit results from using financial status
audit techniques in terms of the amount of adjustments to reported
income; and (4) determine how IRS applied its audit standards, quality
controls, and measurement of audit quality to the use of financial
status techniques.

GAO noted that: (1) on the basis of its review of samples of IRS audits
completed before and after IRS reemphasized the use of financial status
techniques, GAO found no statistically significant change in the
frequency with which these techniques were used or in the types of
returns for which the techniques were used; (2) during both periods,
over 75 percent of the audits using financial status techniques involved
individual returns with business or farm income--the types of taxpayers
that IRS has historically found to be the most likely to underreport
income; (3) financial status audit techniques vary in the need for
taxpayer contact and how much additional burden or intrusiveness may be
perceived by the taxpayer; (4) financial status audits have been
criticized by tax professionals and others for, among other things,
seeking information about financial status without having evidence of
unreported income; (5) such intrusions into taxpayers' spending patterns
could occur before the initial interview and during the initial
interview; (6) IRS used the Personal Living Expense (PLE) form to
inquire about expenses at the time of the notification letter in fewer
than 5 percent of the audits for both the 1992 and 1993 and 1995 and
1996 periods; (7) the case files showed that auditors infrequently asked
intrusive, financial status type questions at the initial interview; (8)
concerning the results, auditors made no adjustments to the individual's
reported income attributable to the use of financial status audit
techniques in 83 percent of the audits in which these techniques were
used; (9) IRS has three tools to oversee the use of financial status
audit techniques: (a) audit standards to guide auditors; (b) supervisory
review of auditors' adherence to the standards; and (c) a system to
measure adherence to the standards; (10) while these tools offered
important controls over the use of the financial status techniques, they
each have limitations; and (11) on the basis of GAO's review of IRS
audit workpapers, the lack of specific criteria may have contributed to
the relatively large percentage of audits in which the use of financial
status audit techniques resulted in no adjustments to income.

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

 REPORTNUM:  GGD-98-38
     TITLE:  Tax Administration: More Criteria Needed on IRS' Use of 
             Financial Status Audit Techniques
      DATE:  12/30/97
   SUBJECT:  Tax return audits
             Tax nonpayment
             Statistical methods
             Tax returns
             Income taxes
             Taxpayers
             Auditing procedures
             Auditing standards
             Tax administration systems
IDENTIFIER:  IRS Audit Information Management System
             IRS Examination Quality Measurement System
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


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


Report to the Chairman, Committee on Ways and Means, House of
Representatives

December 1997

TAX ADMINISTRATION - MORE CRITERIA
NEEDED ON IRS' USE OF FINANCIAL
STATUS AUDIT TECHNIQUES

GAO/GGD-98-38

Financial Status Audit Techniques

(268762)


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

  AICPA - American Institute of Certified Public Accountants
  AIMS - Audit Information Management System
  DCI - Data Collection Instrument
  EQMS - Examination Quality Measurement System
  IRS - Internal Revenue Service
  PLE - Personal Living Expense

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


B-275099

December 30, 1997

The Honorable Bill Archer
Chairman, Committee on Ways and Means

The Honorable Nancy L.  Johnson
Chairman, Subcommittee on Oversight
Committee on Ways and Means
House of Representatives

Each year, Internal Revenue Service (IRS) auditors identify billions
of dollars in additional income taxes owed through audits of
individual taxpayers.  Such tax audits have been a fundamental part
of IRS' enforcement strategy for many years, helping to ensure that
taxpayers pay the amount of taxes they owe.  Tax audits may occur in
a variety of forms, ranging from a simple review of a return with
little taxpayer contact to a detailed on-site examination and
investigation of a taxpayer's financial records.  Increasingly,
however, the way IRS conducts its audits has been criticized by
taxpayers, tax professionals, and Congress as being overly intrusive
and burdensome for taxpayers. 

Much of this criticism has stemmed from IRS' reemphasis on detecting
unreported income.  In the early 1990s, IRS managers became concerned
that auditors were not fully using audit techniques designed to
identify unreported income.  In a 1994 initiative, to address this
concern, IRS implemented a training program to reemphasize the need
to consider a taxpayer's financial status by focusing on whether the
taxpayer's income and expenses were roughly proportional.  The
training program reemphasized certain audit techniques for
identifying unreported income.  These techniques are sometimes
referred to as financial status audit techniques. 

You asked that we review IRS' use of financial status audit
techniques.  In this report, we (1) estimate how frequently IRS used
financial status audit techniques in audits closed in tax years prior
to the 1994 initiative (1992 and 1993) and in tax years following the
initiative (1995 and 1996); (2) consider how IRS' need to contact
taxpayers for additional taxpayer information when using financial
status techniques might intrude on taxpayers; (3) estimate the audit
results from using financial status audit techniques in terms of the
amount of adjustments to reported income; and (4) determine how IRS
applied its audit standards, quality controls, and measurement of
audit quality to the use of financial status techniques. 

As our primary method for addressing these objectives, we selected
random samples of audits of individual returns completed before and
after IRS implemented its financial status audit initiative in 1994
and examined the IRS audit workpapers for the sampled tax returns. 
We discussed our observations about the audits we reviewed with IRS
officials.  We did not contact individual taxpayers about the audits
we reviewed.  However, we discussed the issues of intrusiveness and
burdens of IRS' financial status audits with knowledgeable tax
professionals. 


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

IRS defines the tax gap as the amount of tax that taxpayers owed but
have not paid.  IRS estimates the individual income tax gap to be
$95.3 billion for 1992.  Unreported income accounts for a major
portion of this tax gap--$58.6 billion or over 60 percent.  In the
early 1990s, IRS became concerned that its auditors were not fully
probing for income that should have been, but was not, reported on
tax returns.  This concern as well as others led IRS to reemphasize
the need for its auditors to consider a taxpayer's financial status
and to probe for unreported income.\1 This reemphasis came to be
known as the financial status audit program.\2

IRS initiated the financial status audit program in late 1994 with a
training course for auditors.\3 In the training course, IRS stressed
the importance of identifying unreported income by determining
whether the taxpayer's reported income roughly conforms to his or her
spending.  Such an evaluation requires consideration of the
taxpayer's spending patterns in addition to verifying items reported
on tax returns.  If reported income and spending patterns differ, the
auditor is supposed to decide whether the difference is significant
enough to warrant asking the taxpayer for an explanation. 

The training course stressed the importance of meeting with
taxpayers, checking nontraditional data sources (such as state and
local governments), and using four indirect audit techniques.\4 These
four techniques, the cornerstones of financial status audits, are

  -- bank deposit analysis, in which the auditor uses the taxpayer's
     bank statements to ensure that total deposits are accounted for
     on the tax return or as nontaxable receipts;

  -- net worth method, in which the auditor analyzes changes in the
     taxpayer's assets to determine any potential for unreported
     income;

  -- normal markup/unit of sales method, in which the auditor uses
     the taxpayer's cost of goods sold and average markups within the
     industry to estimate business gross receipts; and

  -- cash transaction (Cash-T) method, in which the auditor compares
     the taxpayer's expenditures to income sources.  Under this
     method, if a taxpayer's expenditures exceed reported income and
     the source for such expenditures cannot be explained, the excess
     represents potential unreported income. 

The Cash-T method also includes a preliminary Cash-T in which the
auditors use only the information available on the tax return to
determine whether the expenditures exceeded reported income.  The
preliminary Cash-T can be completed without contacting the taxpayer
for information. 

The consideration of a taxpayer's financial status and the use of
these techniques to probe for unreported income are not new concepts. 
Historically, the techniques have been used in fraud and criminal
investigation cases, but they have also been available for use by
other IRS auditors.  IRS officials noted that the use of financial
status techniques has been mentioned in the Internal Revenue Manual
at least as far back as 1961.  According to IRS officials, the 1994
financial status initiative was intended primarily to reemphasize
instructions that auditors receive in other IRS training courses. 

By early 1995, IRS was receiving considerable criticism about audits
using these financial status techniques.  The American Institute of
Certified Public Accountants (AICPA), Members of Congress, and
various taxpayer groups were concerned that these audits were more
time consuming and intrusive than other auditing techniques.\5 AICPA
officials had several concerns about the taxpayer burden and
intrusiveness that they associated with IRS' use of financial status
techniques.  Specifically, they were concerned about IRS' practice of
asking financial status questions at the initial interview before
having any evidence of underreported income.  Similarly, AICPA
officials were concerned about IRS sending a request for personal
living expense (PLE) information with the letter notifying the
taxpayer of the audit, before finding any evidence of unreported
income. 

In response to these criticisms, IRS provided additional instructions
to its auditors to clarify the intent of financial status audits. 
Between August 1995 and March 1996, three memoranda were issued from
the Office of the Assistant Commissioner (Examination) to Regional
Chief Compliance Officers to provide the clarifications.  The August
memorandum supported the use of financial status techniques but urged
auditors to use sound judgment in asking financial status questions
at the initial interview, particularly when no indication of
underreported income existed.  The December 1995 and March 1996
memoranda provided similar instructions, including guidance
indicating that PLE forms should not automatically accompany
notification letters.  AICPA officials acknowledged to us that these
instructions helped to reduce some of their concerns, but they said
they were still concerned about the added time and intrusiveness
associated with IRS' use of financial status audit techniques. 


--------------------
\1 IRS had also announced a plan in 1994 to conduct an expanded
program of audits of a stratified random sample of returns to measure
taxpayer compliance.  IRS wanted to ensure that these compliance
audits were accurate because the results would be used to update the
estimates of the income tax gap and the formulas used to objectively
select tax returns for future audits.  Because of public and
congressional concerns about the scope and intrusiveness of this
proposed program and IRS' budget constraints, it was ultimately
cancelled. 

\2 When originally conceived, the financial status audit program was
referred to as the Economic Reality Program. 

\3 The term auditor, as referred to in this report, includes revenue
agents and tax auditors because both do face-to-face audits with
taxpayers.  For individuals, revenue agents usually audit taxpayers
who report significant amounts of business income or file very
complex returns while tax auditors usually audit those who do not
report significant amounts of income or file simpler returns. 

\4 In its letter commenting on a draft of this report, IRS said that
it uses the term "financial status audit" to mean nothing more than
an examination of a return where a potential unreported income issue
was identified through the analysis of the taxpayer's financial
status.  IRS auditors then use indirect techniques to check out this
potential for unreported income.  We refer to these indirect
techniques as financial status techniques. 

\5 Auditors generally have three sources for verifying the taxpayers'
income:  (1) taxpayers (e.g., records, admissions); (2) third-parties
(e.g., those filing information returns with IRS to report payments
made to taxpayers); and (3) one or more of the financial status audit
techniques. 


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

On the basis of our review of samples of IRS audits completed before
and after IRS reemphasized the use of financial status techniques, we
found no statistically significant change in the frequency with which
these techniques were used or in the types of returns for which the
techniques were used.\6 For audits completed in 1992 and 1993, before
IRS' reemphasis on financial status techniques, we estimated that
auditors used the techniques on about 24 percent of the universe of
556,000 audits.  Similarly, for audits completed in 1995 and 1996,
after the reemphasis, we estimated that auditors used the techniques
on about 22 percent of the universe of 421,000 audits.  Because IRS
lacks specific criteria on when to use the techniques, we could not
determine whether the frequency of use was appropriate for either
time period. 

We also estimated that, during both periods, over 75 percent of the
audits using financial status techniques involved individual returns
with business or farm income--the types of taxpayers that IRS has
historically found to be the most likely to underreport income. 
Virtually all of the audits from both periods that used one or more
of these techniques used Cash-Ts (preliminary or comprehensive), bank
deposit analyses, or both. 

Financial status audit techniques vary in the need for taxpayer
contact and how much additional burden or intrusiveness may be
perceived by the taxpayer.  For example, IRS auditors used only a
preliminary Cash-T in about 23 percent of the 1995 and 1996 audits we
reviewed where a financial status audit technique was used.  This
technique imposed no additional burden on the taxpayer because it
requires no contact with the taxpayer.  In the remaining 77 percent
of the audits, some additional contact with the taxpayer was
necessary to obtain financial status information.  We did not attempt
to measure the additional burden or intrusiveness attributable to the
use of financial status techniques in these cases because first, IRS
has no definitions of burden and intrusiveness and, second, even if
it had the definitions, the audit workpapers did not contain
sufficient quantitative data for such measurements. 

We were able to examine at least two points where intrusiveness could
occur, however.  Financial status audits have been criticized by tax
professionals and others for, among other things, seeking information
about financial status without having evidence of unreported income. 
Such intrusions into taxpayers' spending patterns could occur (1)
before the initial interview and (2) during the initial interview. 
Critics suggested that such intrusions increased after the 1994
initiative.  Our analysis of the sampled audits for the two time
periods indicates that the frequency of occurrence of these two
alleged types of intrusions have not changed much. 

First, IRS used the PLE form to inquire about expenses at the time of
the notification letter in fewer than 5 percent of the audits for
both the 1992 and 1993 and 1995 and 1996 periods.  Second, the case
files showed that auditors infrequently asked intrusive, financial
status type questions at the initial interview.  Of the 16 questions
identified by AICPA as questions it considered intrusive, most were
asked in the initial interview in fewer than 5 percent of the audits. 
The frequencies with which the questions were asked were about the
same for both periods. 

Concerning the results, auditors made no adjustments to the
individual's reported income attributable to the use of financial
status audit techniques in 83 percent of the audits in which these
techniques were used.  Notwithstanding this relatively high no-change
rate, the use of financial status audit techniques helped to identify
a significant amount of unreported income in some audits.  On the
basis of our sample for the 1995 and 1996 period, we estimated that
IRS was able to identify over $300 million in underreported income
using the financial status techniques. 

IRS has three tools to oversee the use of financial status audit
techniques:  (1) audit standards to guide auditors, (2) supervisory
review of auditors' adherence to the standards, and (3) a system to
measure adherence to the standards.  Our analyses focused on how IRS
applied these tools to the use of financial status audit techniques. 
While these tools offered important controls over the use of the
financial status techniques, they each have limitations.  For
example: 

  -- IRS' nine audit standards did not have specific criteria to
     guide auditors on when to use financial status techniques and to
     what degree. 

  -- Our analysis of the IRS workpapers indicated that supervisory
     review of the audits appeared to be limited.  IRS officials we
     met with acknowledged that managers cannot review all audits,
     and the managers told us they tried to at least maintain general
     oversight of auditors' ongoing audit inventories. 

  -- Because the standards did not include specific criteria on when
     to use financial status audit techniques, IRS' measurement did
     not address whether the auditors should have used the
     techniques. 

As the administrator of the nation's tax system, IRS is responsible
for identifying the correct amount of tax that is owed.  Because our
tax system is based on voluntary compliance, an appropriate balance
must be maintained between collecting evidence and information to
assist the auditor in identifying the correct tax and avoiding
unnecessary burden and intrusiveness for the large majority of
taxpayers.  More specific criteria for IRS auditors to use in making
case-by-case decisions about whether and to what extent to use
financial status audit techniques would be helpful to auditors in
achieving that balance.  On the basis of our review of IRS audit
workpapers, we believe that the lack of specific criteria may have
contributed to the relatively large percentage of audits in which the
use of financial status audit techniques resulted in no adjustments
to income.  During the course of our work, IRS agreed that it needs
more specific criteria to guide its auditors in exercising their
judgment to use the financial status techniques. 


--------------------
\6 We selected two samples, totaling 838 audits.  The before sample
contained 484 completed audits, and the after sample contained 354
audits.  A more complete description of our sampling methodology can
be found in appendix I. 


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

To determine the extent to which IRS' use of financial status
techniques has changed, we selected random samples of audits of
individual returns completed before and after IRS began reemphasizing
the techniques in 1994.  We selected these samples from IRS' Audit
Information Management System (AIMS) database.\7

For the "before" sample, we selected audits that were opened on
individuals from October 1991 through October 1992 and closed during
fiscal years 1992 and 1993.  For the "after" sample, we selected
audits that were opened from October 1994 through October 1995 and
closed during fiscal years 1995 and 1996.\8 Each sample audit
included one or more individual income tax returns.  Our sample
contained 838 valid audits selected from an estimated population of
977,000 audits.  All the numbers used in this report are estimates
developed on the basis of weights assigned to the sampled audits so
that they represent the population from which we sampled.  See
appendix I for a more detailed description of our sampling
methodology and the procedures used to develop our estimates. 

We used the IRS workpapers associated with each audit to determine
whether and how auditors used the financial status techniques and
which type of techniques were used.  For each sample audit, using a
data collection instrument that we developed, we gathered specific
information from the case files about the types of techniques used,
amounts of any adjustments to taxable income and tax liability, types
of questions asked the taxpayers, and information about both the
auditor and taxpayer.  We also met with National Office officials
responsible for implementing the financial status program to discuss
our sampling methodology and results.  We did not determine whether
IRS' auditors made appropriate choices in deciding when to use
financial status techniques and what techniques to use because IRS
had no specific criteria against which to make this judgment. 

To obtain information on how the use of financial status techniques
increased the need for taxpayer contact and might have affected the
taxpayer, we again used data from the audit workpapers.  We collected
information from the case files on the types of techniques being
used, whether or not Cash-Ts were preliminary or comprehensive, the
nature of the taxpayer contacts, the types of questions asked at
initial interviews, and whether or not IRS requested PLE information
when first notifying the taxpayer of the audit.  Additionally, we met
with IRS' National and Field office officials to learn how each
technique was used.  As part of our work on this issue, we discussed
the financial status program with officials at AICPA.  These
officials raised several concerns about IRS' use of financial status
techniques and the whole approach to audits resulting from the
emphasis on the techniques.  To the extent possible, we used our
sample data to evaluate these concerns. 

To determine the results of audits using financial status techniques,
we used the samples and workpapers previously discussed.  For each
audit, we recorded the adjustments to income and additional taxes
found on all returns.  We also recorded the amount of the changes to
income attributable to the use of one or more of the financial status
techniques. 

To determine how IRS applied its audit standards, quality controls,
and quality measurement to the use of financial status techniques, we
met with officials in the Examination Division, including the Quality
Measurement staff, at the National Office and four district
offices.\9 We also discussed quality review procedures with group
managers at the district offices.  We obtained copies of the audit
standards and reviewed their applicability to the financial status
program.  Otherwise, we did not evaluate the adequacy of the
standards.  We reviewed IRS' Examination Quality Measurement System
(EQMS) to determine how IRS measures audit quality and what the
measures show.  At three of the four district offices, we examined
several EQMS cases, selected by IRS personnel, to see how EQMS
reviews were done.  We did not examine any cases at the Philadelphia
district office because all EQMS reviews in that region are done at
another district office.  We did not try to assess the accuracy of
EQMS reviews.  (See appendix II for a summary of IRS' audit
standards.)

We requested comments on a draft of this report from the Commissioner
of Internal Revenue.  On November 20, 1997, we received written
comments from IRS, which are summarized at the end of this letter and
are reproduced in appendix IV.  These comments have been incorporated
into the report where appropriate. 

We performed our audit at IRS headquarters offices in Washington,
D.C., and at district offices and service centers in Fresno and
Oakland, CA; Baltimore, MD; Philadelphia, PA; and Richmond, VA.  Our
work was done between October 1996 and August 1997 in accordance with
generally accepted government auditing standards. 


--------------------
\7 We did not validate the accuracy of information in the AIMS
database. 

\8 Our sample excluded audits closed on AIMS that did not involve
looking at taxpayers' books and records and those that could not be
expected to use the financial status techniques.  These audits
included correspondence audits at IRS Service Centers and limited
scope audits to pass through adjustments from partnership audits to
the partners, identify nonfilers, prepare substitute returns for
nonfilers, and review taxpayers' claims for refund. 

\9 These district offices, selected for their proximity to our staff
locations, were Baltimore, Oakland, Philadelphia, and Richmond. 


   LITTLE CHANGE SHOWN IN THE USE
   OF FINANCIAL STATUS TECHNIQUES
------------------------------------------------------------ Letter :4

IRS' renewed emphasis on financial status audit techniques produced
little, if any, change in how often these techniques were used. 
Comparing audits done before and after IRS' emphasis on financial
status, we estimated that the use of one or more of the financial
status techniques was 24 percent for the 1992 and 1993 period and 22
percent for the 1995 and 1996 period.  The difference in these
percentages is not statistically significant.  During both periods,
financial status techniques were used predominately on returns
involving business or farm income.  IRS research has found that
taxpayers with these types of income are more likely to underreport
income than taxpayers whose income is reported by third parties on
information returns.  Table 1 compares the two periods we reviewed. 



                                Table 1
                
                 Number and Percentages of Audits That
                Used Financial Status Techniques, Before
                       and After IRS' Reemphasis

                                                Sampling period
                                          ----------------------------
Audits                                        1992-1993      1995-1996
----------------------------------------  -------------  -------------
Estimated total audits\a                        556,000        421,000
Percentages of audits that used one or              24%            22%
 more financial status technique
Percentage of audits that used one or               75%            84%
 more financial status technique
 involving returns with business or farm
 income
Percentage of audits with business or               38%            34%
 farm income where one or more of the
 financial status techniques were used
Percentage of audits with no business or            12%             7%
 farm income where one or more of the
 financial status techniques were used
----------------------------------------------------------------------
\a This estimate was calculated on the basis of our sample of audits
and is adjusted to account for audits that were excluded or missing. 
(See app.  I for additional information on the sampling methodology.)

Source:  GAO analysis of IRS audit workpapers. 

IRS managers were concerned that auditors were not making use of
techniques to identify unreported income.  The financial status
program and the associated training was designed to correct this
problem.  IRS officials could not tell us why the percent of
financial status audits had not changed after the reemphasis and
training.  However, they noted that one reason may have been because
of the limited amount of follow-up training provided by the districts
and the limited amount of National Office oversight due to IRS'
reorganization activities after the initial training.  In commenting
on our draft report, IRS officials indicated that the financial
status training focused less on increasing the use of a specific
technique and more on improving the auditors' ability to identify
unreported income. 

We also analyzed whether IRS changed the types of techniques being
used.  We found no significant change in usage by type of financial
status technique since the reemphasis.  Generally, only two
techniques were used, often in combination, during our two sample
periods.  Table 2 describes the results of this analysis. 



                                Table 2
                
                   Types and Percentages of Financial
                Status Techniques Used in Audits Before
                       and After IRS' Reemphasis

                                                Sampling period
                                          ----------------------------
Audits using techniques                     1992-1993\a    1995-1996\a
----------------------------------------  -------------  -------------
Estimated number of audits                      136,000         91,000
Audits using Cash-T (preliminary and                50%            47%
 comprehensive) only
Audits using bank deposit analysis only             29%            21%
Audits using both Cash-T and bank                   20%            32%
 deposit analysis
All other combinations                               \b             1%
======================================================================
Total                                              100%           100%
----------------------------------------------------------------------
\a Percentages may not total to 100 due to rounding. 

\b Percentage equals less than 1/10th of 1 percent. 

Source:  GAO analysis of IRS audit workpapers. 

To put the data presented in tables 1 and 2 in perspective, in 1995,
about 116 million taxpayers filed their 1994 individual income tax
returns.  On the basis of historical data and information from our
sample, we estimate that between 126,000 and 183,000 will receive an
audit that uses at least one of the four financial status techniques
during the 3 years before the statute of limitations expires. 


   NEED FOR TAXPAYER CONTACT WHEN
   USING FINANCIAL STATUS
   TECHNIQUES VARIES
------------------------------------------------------------ Letter :5

Financial status audit techniques vary in the extent of additional
taxpayer contact needed and the amount of information being sought
from taxpayers.  IRS has no data showing how much additional taxpayer
contact is associated with each technique or how intrusive the
additional information needed might be.  However, we were able to
make some general observations based on our review of the workpapers. 

For example, the Cash-T method can be separated into two
types--preliminary and comprehensive.  In the preliminary Cash-T, the
auditor uses only information available on the tax return to identify
any indications of unreported income.  This technique, therefore,
requires no additional contact with the taxpayer.  Of the estimated
126,000 to 183,000 audits of tax year 1994 individual returns in
which IRS used a financial status technique, we estimated that
between 29,000 and 42,000 of these audits (23 percent) only used a
preliminary Cash-T, requiring no response from the taxpayer. 

The comprehensive Cash-T and each of the other techniques require
some additional taxpayer contact.  The amount of contact required and
information sought can vary with each taxpayer and the type of
financial status technique used.  In a comprehensive Cash-T, the
auditor needs information from the taxpayer on nonreturn items such
as cash on hand, savings, and PLE.  For a bank deposit analysis, the
auditor requires access to the taxpayer's bank account records and
may require considerable taxpayer contact to ask the taxpayer to
explain significant discrepancies between total deposits and the
income shown on the tax return.  The net worth and normal markup
methods require taxpayer contact primarily to explain any identified
discrepancies. 

AICPA has been among the critics of IRS' reemphasis on financial
status audits since the program began in late 1994, claiming that IRS
auditors use the techniques without having any evidence that
taxpayers have underreported income.  Such intrusions into taxpayer's
spending patterns could occur at two points--(1) before the initial
interview and (2) during the initial interview.  Critics suggested
that such intrusions increased after the 1994 initiative.  Using the
data gathered from our reviews of IRS' audit workpapers, we looked at
the frequency of the two concerns. 

We gathered information on how often IRS used the initial
notification letter to request that the taxpayer provide PLE
information.  We found no significant difference between the 1992 and
1993 period (before the reemphasis on financial status) and the 1995
and 1996 period (after the reemphasis).  During both periods, less
than 5 percent of the initial notification letters to the taxpayers
also requested that they provide information on their PLE. 

Recognizing the potential for intrusiveness, the Acting Assistant
Commissioner (Examination) in a March 1996 memo, clarified the PLE
instructions.  The memo indicated that while auditors had the
responsibility to secure an overall financial picture of the
taxpayer, they were not expected to automatically request PLE
information with the notification letter.  According to AICPA
officials, sending PLE forms with the notification letters has
decreased since the distribution of this memo. 

We also gathered information on the types of questions IRS auditors
asked taxpayers at opening interviews.  Financial status critics
believe that questions designed to determine the taxpayer's financial
status were inappropriate unless IRS had evidence that the taxpayer
had underreported income.  AICPA officials provided a list of the
questions, which focused on personal spending habits such as how
often a taxpayer eats at restaurants and where a taxpayer vacations. 

Based on our analysis of the documents in the case files, most of
these interview questions occurred in fewer than 5 percent of the
audits.  For the 1995 and 1996 sample period, only four of the
questions were asked during the initial interview in over 10 percent
of the audits.\10 In addition, the frequency in which the questions
were asked was about the same in our samples of audits for 1992 and
1993 and for 1995 and 1996.  Appendix III provides information about
the specific questions and how often they were asked. 


--------------------
\10 The frequency for these four questions ranged from 11 percent to
24 percent of the audits. 


   RESULTS FROM AUDITS USING
   FINANCIAL STATUS TECHNIQUES
   HAVE BEEN MIXED
------------------------------------------------------------ Letter :6

The results of using financial status techniques have been mixed. 
The use of the techniques resulted in IRS auditors identifying large
amounts of unreported income in some cases.\11 At the same time, a
high percentage of audits resulted in no adjustments to reported
income attributable to the use of financial status techniques.\12
Table 3 summarizes these results. 



                                Table 3
                
                  Estimated Results of Using Financial
                 Status Techniques, 1992-1993 and 1995-
                                  1996

                                                Sampling period
                                          ----------------------------
Audits using techniques                       1992-1993      1995-1996
----------------------------------------  -------------  -------------
Estimated number of audits                      136,100         91,400
Use of financial status techniques                  81%            83%
 resulted in no adjustment to reported
 income
Use of financial status techniques                  12%           8%\a
 resulted in adjustments to reported
 income of less than $10,000
Use of financial status techniques                   7%             9%
 resulted in adjustments to reported
 income of $10,000 or over
Total                                              100%           100%
----------------------------------------------------------------------
\a This figure includes one case in which the use of financial status
techniques actually resulted in reducing the taxpayer's reported
income. 

Source:  GAO analysis of IRS audit workpapers. 

IRS reemphasized the use of financial status techniques to address
its concerns with finding unreported income.  In the audits we
reviewed in our 1995 and 1996 sample, we estimated that auditors used
financial status techniques to identify unreported income totaling
over $300 million.\13 Our review of the IRS workpapers indicated that
the auditors were unlikely to have identified unreported income
without using the techniques.  The workpapers did not show that this
income was reported on an information return or identified by the
taxpayer, the other two primary techniques used to verify the
accuracy of reported income. 

However, table 3 shows that the use of financial status techniques
has resulted in no adjustments to income in a significant number of
cases.  For example, in our 1992 and 1993 sample, 81 percent of the
audits using financial status techniques resulted in no adjustments
to reported income attributable specifically to the techniques. 
Similarly, for the 1995 and 1996 sample, 83 percent resulted in no
adjustment to reported income attributable to the use of the
techniques.\14

Audits having no change attributable to the use of financial status
audit techniques may have had changes attributable to other audit
techniques.  These no-change audits were closed with either (1) no
changes to any tax issue or (2) changes such as reducing claims for a
tax deduction, exemption, or credit after the auditor reviewed the
taxpayer's documentation.  For the 1992 and 1993 audits having an 81
percent no-change rate, 23 percent had no change for any reason and
58 percent had changes to taxable income that were not attributable
to the use of financial status techniques.  For 1995 and 1996, the 83
percent no-change rate breaks out as 28 percent with no change for
any reason and 55 percent with changes to taxable income that were
not attributable to the use of financial status techniques.\15

This high percentage of no change attributable to the use of
financial status techniques raises issues about whether IRS can
further help auditors in judging when and how to use these
techniques.  Given the complexity of the tax code and the fact that
tax return forms provide for limited, if any, explanation of the
numbers entered by the taxpayer, it is not reasonable to expect an
adjustment every time a financial status technique is used nor is it
desirable that all auditor judgment be removed from the decision
about when to use the techniques.  It is important, however, that IRS
make the most effective and efficient use of its limited resources
while striking an appropriate balance between collecting information
and evidence to assist the auditor in identifying the correct tax,
and avoiding unnecessary burden and intrusiveness for the taxpayers. 
Thus, the best interest of both IRS and the taxpayers is achieved
when the no-change rate is at some acceptable low point.  To this
end, we believe that more specific criteria on when to use financial
status techniques would provide auditors with additional context
around which to exercise their professional judgment on a
case-by-case basis, and would likely result in a reduced no-change
rate. 


--------------------
\11 Adjustments to taxable income include changes to income, such as
wages or business gross receipts, business expenses, personal
deductions, and exemptions.  Financial status techniques primarily
detect unreported income. 

\12 On average, audits using financial status techniques made larger
adjustments to reported income than those not using the techniques. 
However, the information available in IRS' workpapers did not allow
us to determine whether this difference was attributable to the use
of the techniques or the type of returns on which they were used. 
Rather than compare the results of audits that used and did not use
the techniques, we focused on the results produced by using the
techniques. 

\13 Actual changes in income identified using the techniques for
audits we reviewed ranged from a reduction of about $8,700 to an
increase of about $162,600. 

\14 In commenting on our draft report, IRS said that the no-change
rate attributable to the use of the techniques does not mean that the
usage was inappropriate.  We generally agree but also believe that
the rate we found seems high. 

\15 In its letter commenting on a draft of this report, IRS said that
although use of the techniques may not have led to changes to
reported income, usage could have helped identify other tax changes. 
We did not find this outcome in any of the sampled audits we
analyzed. 


   IRS TOOLS TO OVERSEE USE OF
   FINANCIAL STATUS TECHNIQUES
------------------------------------------------------------ Letter :7

IRS has three primary tools to oversee use of financial status audit
techniques:  (1) audit standards to guide auditors, (2) supervisory
review of auditors' adherence to the standards, and (3) a system to
measure adherence to the standards.  Our analyses focused on how IRS
applied these tools to the use of financial status audit techniques. 
While these tools offered important controls over the use of the
financial status techniques, they each have limitations.  For
example, the audit standards do not guide auditors on when and when
not to use financial status techniques.  IRS' managers at the group
level review a small portion of the audits because of a lack of time
caused by other duties.  IRS' measurement system, like the standards,
focused on whether financial status techniques, when used, were used
correctly from a technical perspective, not on when to use the
techniques and to what degree. 


      AUDIT STANDARDS LACK
      SPECIFIC CRITERIA FOR USING
      FINANCIAL STATUS TECHNIQUES
---------------------------------------------------------- Letter :7.1

IRS uses its audit standards, which have evolved since the 1960s, to
define audit quality.  However, the standards do not offer specific
criteria to guide auditors on when and when not to use financial
status techniques and to what degree.  Instead, the standards focus
on whether actions were taken and, if so, whether they were taken
correctly from a technical perspective. 

IRS uses nine audit standards to address the scope, audit techniques,
technical conclusions, workpaper preparation, reports, and time
management of an audit.  Each standard is composed of key elements
that operationally define a quality examination.  IRS guidance
stipulates that for a standard to be rated as being "met," each of
the key elements must be rated as "met" or "not applicable." The
standards and the associated key elements are summarized in appendix
II. 

Of the nine audit standards, Standard 2, Probes for Unreported
Income, has four key elements that address whether the auditor (1)
considered the adequacy of internal controls, (2) considered the
types of books and records maintained, (3) considered the taxpayer's
financial status, and (4) appropriately used indirect audit
techniques to probe for unreported income.  These last two elements
directly address financial status analyses and audit techniques. 
Under Standard 2, auditors are instructed to consider financial
status in all audits and only use a financial status audit technique
when they suspect unreported income. 

However, IRS did not provide specific criteria in the standards to
help auditors decide when unreported income is likely.  The key
element for evaluating appropriate use of these techniques addressed
whether the auditor considered using a technique, selected the
appropriate technique, and applied it correctly.  Nothing in the
standard provides the auditor with specific criteria to determine
when to use or not use a given technique or to what degree to use it. 
For example, IRS has not instructed auditors on how extensively to
consider a taxpayer's financial status and when that consideration
should prompt the use of a technique to probe for unreported income. 
Nor has IRS instructed auditors on how large a discrepancy between
reported income and expenses should be to justify more in-depth
probing.  On the basis of our review of the audit workpapers, we
believe that this lack of criteria has probably contributed to the
large percentage of audits in which the use of financial status
techniques resulted in no adjustments to income. 

During the course of our work, IRS agreed with us that it needs
specific criteria to better guide its auditors on using the financial
status techniques.  According to an IRS official, sections of the
Internal Revenue Manual are being revised to better instruct tax
auditors and revenue agents about when and when not to use financial
status techniques and to what degree to use them in probing for
unreported income.  In September 1997, we received a draft of the
revised manual sections.  Our initial review of these revised
instructions indicated that they offered some guidance on when to use
financial status techniques but did not provide specific criteria. 
For example, the revisions indicate that if a preliminary analysis
yields a Cash-T that is materially out of balance, the auditor should
use subsequent interviews and information gathering to resolve the
imbalance. 

The instructions define "material imbalance" as the significance of
an item in determining the correct tax liability.  The instructions
require auditors to use their judgment on the return as a whole and
the items that comprise that return.  In using their judgment on
whether the imbalance is material, the auditors must consider such
factors as the comparative and absolute size of the imbalance as well
as the relationship between the size of the imbalance and the tax
liability.  However, IRS has not provided instructions to guide the
auditor when analyzing the comparative or absolute size of the
imbalance or when comparing the relationship of the imbalance to the
tax liability.  In commenting on a draft of this report, IRS
officials said it would be impractical to develop specific
quantitative criteria to define materiality. 

We acknowledge that developing quantitative criteria to cover every
situation is difficult and that auditors' judgment is still an
important element of any audit.  However, we believe that the concept
of "material imbalance" could be made more specific by developing
some quantitative criteria that would use the preliminary Cash-T and
establish thresholds for the factors associated with an imbalance
between reported income and estimates of PLE, such as the comparative
size of any imbalance.  If the preliminary Cash-T indicated that the
income reported on the tax return that was available for PLE was
below the threshold--that is, apparently not sufficient to support
the living expenses indicated--the auditor would be expected to
conduct a more detailed probe for unreported income, potentially
using one or more of the other financial status techniques.  If the
preliminary Cash-T showed the taxpayer's reported income to be above
the threshold--that is, apparently sufficient to support the
estimated PLE--using the other financial status techniques would not
be expected. 

In either case, the auditor could decide to go against the criteria
but would be expected to explain the reasons in the workpapers. 
Developing such criteria would be an on-going task, as changes would
likely occur as IRS gained experience about how well the criteria
were working. 


      SUPERVISORS CANNOT REVIEW
      ALL AUDIT WORKPAPERS
---------------------------------------------------------- Letter :7.2

The primary tool used by IRS to control quality is the review of
audit files by managers of audit groups.  The Internal Revenue Manual
requires supervisory review of cases but is vague on exactly when
review is necessary and how it should be documented. 

According to IRS Examination officials, IRS managers cannot review
all audits.\16

Rather, the managers must rely on the experience and judgment of the
auditors because the manager's audit workload and other duties limit
the time available for review.  Further, these officials said budget
constraints will likely cause the managers' span of control to
increase rather than decrease in the future, resulting in more audits
to oversee.  The analysis of our sample supports IRS' assertions that
not all audits are reviewed by managers.  We found evidence of
supervisory review in about 9 percent and 6 percent of the audits for
1992 and 1993 and 1995 and 1996, respectively. 

In the districts we visited, the managers acknowledged that they can
only review a small portion of all ongoing and closed audits for each
auditor annually because of the reasons cited.  Managers told us they
try to spend more time reviewing the work of the least experienced
auditors.  At a minimum, they said they try to maintain an ongoing
discussion with all auditors about their audit inventories. 


--------------------
\16 However, IRS officials told us that managers should review all
unagreed and trainee audits.  Unagreed audits are those that were
closed without the taxpayer agreeing with the auditor's recommended
adjustments to taxable income or tax liability.  For fiscal year
1996, there were about 86,800 audits that were closed as unagreed. 
Trainee audits are ones selected for purposes of training IRS
auditors.  For fiscal year 1996, there were about 117,100 such audits
selected. 


      EQMS MEASUREMENT IS LIMITED
      BY THE DATA COLLECTED
---------------------------------------------------------- Letter :7.3

IRS conducts post-audit quality measurement through EQMS reviews. 
EQMS is IRS' mechanism for collecting information about the audit
process, changes to that process, the level of audit quality, and the
success of any efforts to improve the process and quality.  The
Office of Compliance Specialization, within IRS' Examination
Division, has responsibility for this program.  This office compiles
and maintains a national database of the quality reviews done at the
district level.  This database can be used to identify trends by
district and nationally.  Of the 800,000 face-to-face audits done by
IRS in fiscal year 1996, EQMS staff reviewed a sample of 12,170
audits to measure quality against the nine audit standards. 
According to IRS officials, this sample provided a statistically
valid basis for measuring audit quality.  EQMS staff reviewed the
12,170 audits to determine whether the auditors met the criteria for
each of the auditing standards.  For fiscal year 1996, the
percentages of audits that were rated as having met the standards
ranged from 38 percent for Standard 9, Time Span/Time Charged, to 95
percent for Standard 5, Findings Supported by Law.  (App.  II
summarizes EQMS results since fiscal year 1992.)

Before fiscal year 1997, IRS did not collect data on the reasons key
elements were not met.  Starting in fiscal year 1997, however, IRS
began collecting these data.  For the first 2 quarters of fiscal year
1997, reviewers looked at 2,904 office audits and 2,859 field
audits.\17 Of these audits, IRS rated 84 percent and 78 percent of
the office and field audits, respectively, as having met (i.e.,
passed) the key element under Standard 2 that involves the
consideration of financial status.  Further, 74 percent and 82
percent of these office and field audits, respectively, were rated as
having met the key element under Standard 2 that involves the
appropriate use of financial status audit techniques.  Table 4
summarizes the EQMS-determined reasons auditors did not meet these
key elements of Standard 2.  For example, the most frequent reasons
cited were that auditors did not (1) provide evidence that they had
evaluated financial status, (2) recognize the need to use one of the
financial status techniques, and (3) correctly compute the financial
status technique. 



                                Table 4
                
                 Reasons Two Key Elements of Standard 2
                  Were Rated as Not Met (Oct. 1, 1996
                         through Mar. 31, 1997)

                                             Audits not meeting key
                                                    element
                                          ----------------------------
Key element                                Office audit    Field audit
----------------------------------------  -------------  -------------
Consideration of financial status\a                 424            584

Reasons for being rated as "not met"
----------------------------------------------------------------------
No evidence of evaluation                         61.2%          48.7%
Standard of living/PLE not considered             25.5%          19.7%
Financial history not considered                   3.9%          11.6%
Potential source of funds not considered           3.0%           3.3%
Accumulation of wealth/assets not                  2.8%           4.6%
 considered
Loans (receipts & payments) not                    1.6%           4.8%
 considered
Significant results considered                     1.1%           2.0%
 insignificant
Business environment not considered                0.9%           5.4%

Appropriate use of financial status                 137            265
 techniques\a

Reasons for being rated as "not met"
----------------------------------------------------------------------
Did not recognize need for techniques             65.1%          32.5%
Technique computed incorrectly                    14.7%          42.7%
Significant results considered                    12.8%          10.2%
 insignificant
Did not use appropriate method                     7.3%          14.6%
----------------------------------------------------------------------
\a Percentages may not total to 100 due to rounding. 

Source:  IRS data. 

Knowing the reasons for not meeting the key element or the standard
can provide insights on when the use of the financial status
techniques would and would not be necessary to identify unreported
income.  However, the reasons identified by IRS, like the criteria in
the audit standard on probing for unreported income, have not
addressed the issue of when and when not to use financial status
techniques and to what degree they should be used.  Without this
information, IRS cannot fully measure the quality of audits involving
financial status techniques. 


--------------------
\17 Traditionally, IRS has conducted two types of face-to-face audits
from its district offices:  (1) field audits, in which an IRS revenue
agent visits an individual taxpayer who has business income or a very
complex return and (2) office audits, in which an individual taxpayer
who has a less complex return visits a tax auditor at an IRS office. 


   CONCLUSIONS
------------------------------------------------------------ Letter :8

IRS auditors have used financial status audit techniques for years to
help identify unreported income.  IRS' renewed emphasis on the use of
these techniques appears to have had little impact on how frequently
auditors used them.  Also, neither the type of technique nor the type
of return on which they are used has changed to any statistically
significant degree. 

IRS has not measured how the use of financial status techniques may
add to the burden and intrusiveness of audits.  Use of the
preliminary Cash-T technique added no burden because this technique
does not require additional taxpayer contact.  Use of the other
financial status techniques require some degree of taxpayer contact. 
The amount of contact and the amount of additional information sought
from the taxpayer, however, can vary with each situation. 

The results of using financial status techniques were mixed.  In a
large majority of such audits, no adjustments to income could be
attributed specifically to the techniques.  While it is not
reasonable to expect unreported income to be found every time these
techniques are used, the current rate of no adjustments seems high. 
However, in the remaining audits, the use of the techniques helped
auditors to find unreported income that probably would not otherwise
have been detected. 

This detection capability and the high frequency of no adjustments to
reported income raises the issues of how to decide when and when not
to use financial status techniques and to what degree they should be
used.  Currently, auditors' judgment primarily dictates these
decisions because IRS does not provide the auditors with specific
guidance for determining whether to use financial status audit
techniques.  While an auditor's judgment is likely to continue to
constitute a significant portion of the decisionmaking process,
guidance, in the form of specific criteria, might help reduce the
frequency in which these techniques are used but do not result in
adjustment to income. 

Similarly, supervisory review of audits to guide the auditors'
performance, a key piece of IRS' quality control system, was limited
by workload constraints and when done, seldom addressed the use of
financial status techniques.  Finally, IRS staff reviewed some closed
audits for quality through EQMS, but like the audit standards, these
reviews did not focus on when and when not to use financial status
techniques and to what degree to use them. 

Without establishing specific criteria to guide the usage of
financial status audit techniques, IRS does not have a good basis for
evaluating the auditors' judgment in choosing to use or not use the
techniques.  We believe that such criteria would help IRS auditors
make their decisions.  Given that our tax system is based on
voluntary compliance, an appropriate balance must be maintained
between collecting information to assist the auditor in identifying
the correct tax and avoiding unnecessary burden and intrusiveness for
the large majority of taxpayers.  More specific criteria to use in
making case-by-case decisions about when and to what extent to use
financial status audit techniques would be helpful to auditors in
achieving that balance.  Developing such criteria, however, would
have to be considered a work in progress, with changes and updates
occurring as needed when auditors and managers become more
experienced with their use.  During the course of our work, IRS
agreed that it needs more specific criteria to guide its auditors in
exercising their judgment to use the financial status techniques and
began developing instructions that include such criteria to be
included in the Internal Revenue Manual. 


   RECOMMENDATIONS
------------------------------------------------------------ Letter :9

To provide better assurance that financial status techniques are not
overly burdensome and intrusive to taxpayers and that the most
productive use is made of limited audit resources, we recommend that
the Commissioner of IRS further pursue efforts to develop more
specific criteria on when and to what extent to use financial status
techniques.  To help develop and refine these criteria, we recommend
that the IRS Commissioner

  -- ensure that these specific criteria on using the techniques are
     reflected in the instructions for interpreting the audit
     standards and the evaluations through EQMS and its reason codes
     of how well audits meet these standards;

  -- monitor the use of financial status techniques under the new
     criteria to identify factors associated with successful and
     unsuccessful usage in terms of when and to what extent to use
     the techniques as well as whether the usage identified
     unreported income and if so, in what amounts; and

  -- use these monitoring results to evaluate whether to make further
     revisions to the criteria on using the techniques or in the
     system by which IRS monitors their use. 


   AGENCY COMMENTS AND OUR
   EVALUATION
----------------------------------------------------------- Letter :10

We obtained comments on a draft of this report at a meeting on
November 12, 1997, with officials who represented IRS.  These
officials included the Chief Compliance Officer, the Assistant
Commissioner for Examination and members of his staff, the National
Director of Compliance Specialization and members of his staff, and a
representative from IRS' Office of Legislative Affairs.  The Deputy
Commissioner also documented these comments in a letter dated
November 20, 1997 (see app.  IV). 

In general, IRS agreed with the substance of our report.  It provided
technical comments to clarify specific sections of the report.  These
comments dealt with issues such as the status and nature of the
instructions being developed on using financial status techniques and
IRS' position on intrusiveness of the techniques and on training.  We
have incorporated these comments into the report where appropriate. 

Concerning the recommendations in our report, IRS agreed with our
overall recommendation on developing more specific criteria to guide
auditors in using financial status techniques and generally agreed
with the three recommendations we made to help with this development. 
IRS officials fully agreed to implement all of our recommendations by
October 1998, as reflected in IRS' letter of November 20, 1997. 


--------------------------------------------------------- Letter :10.1

We are sending copies of this report to the Committee's Ranking
Minority Member, the Chairman and Ranking Minority Member of the
Senate Committee on Finance, various other congressional committees,
the Director of the Office of Management and Budget, the Secretary of
the Treasury, and other interested parties.  We will also make copies
available to others upon request. 

Major contributors to this report are listed in appendix V.  If you
have any questions concerning this report, please contact me at (202)
512-9110. 

Lynda D.  Willis
Director, Tax Policy
 and Administration Issues


STATISTICAL METHODOLOGY FOR
EVALUATING FINANCIAL STATUS AUDIT
TECHNIQUES
=========================================================== Appendix I

This appendix describes the methodology we used to sample Internal
Revenue Service (IRS) audits from 1992 and 1993 and from 1995 and
1996.  We used these samples to quantify the differences in audit
practices before and after IRS began its reemphasis on using the
financial status techniques and to estimate the results of these
audits. 


   STUDY POPULATION
--------------------------------------------------------- Appendix I:1

IRS reemphasized its financial status program late in fiscal year
1994.  To determine whether financial audit practices and results had
changed, we compared audits within IRS's Audit Information Management
System (AIMS) database that were completed before and after the
reemphasis in 1994.  We restricted our study population to audits of
books and records that IRS conducted at district offices.  This meant
that we excluded limited-scope audits initiated solely to assess an
additional tax, resulting from an audit of a partnership or
corporation, audits opened as part of IRS' nonfiler compliance
initiative, audits of taxpayer claims, and substitutes for returns in
which IRS prepares a return for a nonfiler.  We expected that
financial status techniques would have the potential to be used on
the audits we included. 

To identify audits that were completed before auditors were exposed
to the emphasis on financial status, we restricted the pre-1994 study
population to the estimated 566,268 audits that had begun in the
period from October 1, 1991, to October 31, 1992, and were completed
by September 30, 1993.  To identify the most current audits
subsequent to the emphasis on financial status, we restricted the
post-fiscal year 1994 study population to the estimated 421,039
audits that had begun in the period from October 1, 1994, to October
31, 1995, and were completed by September 30, 1996.  We selected a
probability sample of audited tax returns from each of the two time
periods.  We then obtained information about the audits by reviewing
IRS's workpapers. 


   SAMPLE SELECTION AND WEIGHTING
--------------------------------------------------------- Appendix I:2

To obtain the sample of audits of books and records, we selected a
stratified, probability sample of 1,232 tax returns from among all
returns audited in district offices by revenue agents and tax
auditors within the fiscal years 1992, 1993, 1995, and 1996 study
periods.  The samples were drawn for 1992 and 1993 and for 1995 and
1996.  The audit associated with each selected tax return included
all returns of a taxpayer that had been completed during the study
periods.  As two of the sampled returns were associated with the same
audit, the initial sample of 1,232 returns resulted in a sample of
1,231 audits.  These returns were stratified by year, income, and
type of return as shown in table I.1. 



                                    Table I.1
                     
                     Distribution of Tax Returns in the AIMS
                           Database by Sample Strata\a

                                            Type of return
                                  ----------------------------------
                  FY                                High
            opened\b   FY closed  Low income      income    Business       Total
--------  ----------  ----------  ----------  ----------  ----------  ==========
In AIMS         1992        1992     235,949      99,297      67,258     402,504
 database
                1992        1993     214,433      96,636      68,386     379,455
                1995   1995-1996     342,897     153,252     105,526     601,675
================================================================================
Total                                                                  1,382,634
 returns
In our          1992        1992          75          75         125         275
 sample
 of AIMS
 databas
 e
                1992        1993         140         144         148         432
                1995   1995-1996         150         150         225         525
================================================================================
Total                                                                      1,232
 returns
--------------------------------------------------------------------------------
\a The low-income returns are nonbusiness returns on which the
taxpayer reported less than $50,000 in income.  The high-income
returns are nonbusiness returns on which the taxpayer reported
$50,000 or more in income.  The business returns are those for which
more than 50 percent of the total income comes from the taxpayer's
farm or sole proprietor business. 

\b The opening year includes the standard fiscal year that begins on
October 1 as well as the first month of the next fiscal year, ending
on October 31. 

Source:  IRS' AIMS database and GAO sampling data. 

The division of the population and sample of audits between different
types of returns is shown in Table I.2.  The low income, high income,
and business columns contain audits associated with one or more
returns from a single sample strata.  The mixed category contains the
audits that included returns from more than one of the tax-return
strata.  Table I.2 also indicates that IRS could not locate IRS audit
workpapers for the 187 audits and that of the 1,044 audits for which
workpapers were located, 838 were eligible for our study because they
were books and records audits.  The final sample for our analyses of
these audits in this report are the 838 audits identified in the next
to last row of table I.2. 



                                                                      Table I.2
                                                       
                                                          Distribution of Audits by Year and
                                                                  Sample Disposition

                          Type of return (opened 1992)                            Type of return (opened 1995)
                      ------------------------------------  -------------------------------------------------------------------------
                         Low    High  Busine  Mixe
Variable              income  income      ss     d   Total    Low income     High income        Business          Mixed         Total     Grand Total
--------------------  ------  ------  ------  ----  ======  ------------  --------------  --------------  -------------  ============  ==============
Audits in AIMS database
-----------------------------------------------------------------------------------------------------------------------------------------------------
All audits            364,82  154,44  95,946  14,2  629,43       265,426         120,403          76,607         11,780       474,216       1,103,650
                           0       3            25       4
Audits in study       342,91  124,05  81,711  6,99  555,67       243,807         100,912          69,573          6,747       421,039         976,715
 population\a              6       3             6       6

Audits in our sample
-----------------------------------------------------------------------------------------------------------------------------------------------------
=====================================================================================================================================================
Total                    207     204     250    46     707           144             139             213             28           524           1,231
Workpapers not found      25      27      31     6      89            26              25              39              8            98             187
 by IRS
Workpapers available     182     177     219    40     618           118             114             174             20           426           1,044
Excluded audits\b         15      46      54    19     134            13              24              27              8            72             206
Audits eligible for      167     131     165    21     484           105              90             147             12           354             838
 study\c
Percent available        88%     87%     88%   87%                   82%             82%             82%            71%                           85%
-----------------------------------------------------------------------------------------------------------------------------------------------------
\a Estimated books and records audits. 

\b Excluded audits included correspondence audits at IRS Service
Centers and limited scope audits to pass through adjustments from
partnership audits to the partners, identified nonfilers, prepared
substitute returns for nonfilers, and reviewed taxpayers' claims for
refund. 

\c Books and records audit. 

Source:  IRS' AIMS database and GAO sampling data. 

The items in the AIMS database that served as our sampling frame are
individual tax returns, not audits.  Because an audit can include
multiple tax returns, the effect of multiple returns has been
incorporated in the weighting of the sampled audits in the analysis. 
The weights and sampling errors have been calculated using a
multiplicity estimator in which each sampled audit is weighted to
account for the total number of associated returns in the AIMS
sampling frame.\1


--------------------
\1 Sirken, Monroe G.S.:  Stratified Sample Surveys with Multiplicity. 
Journal of the American Statistical Association, March 1972, Vol. 
67, pp.  224-227. 


   SAMPLING ERRORS AND CONFIDENCE
   INTERVALS OF ESTIMATES
--------------------------------------------------------- Appendix I:3

The results shown in this report are estimates because they are based
on the sample of audits drawn from the total population of all
eligible audits.  The accuracy of these estimates is quantified by
their sampling errors, expressed as 95-percent confidence intervals. 
In table I.3, for example, the estimate that 24 percent of the 1992
audits used a financial status audit technique is surrounded by a
confidence interval of + 5 percentage points, indicating that we are
95 percent confident that the actual percentage in the population of
all audits lies between 19 and 29 percent.  The comparison column of
the same table indicates that the difference of 3 percent between the
1992 and 1995 samples is surrounded by a 95-percent confidence
interval of + 6 percentage points, indicating that we are 95 percent
confident that the difference between the 1992 and 1995 audits lies
between -3 and +9 percent.  Since, in this instance the 95-percent
confidence interval included the possibility that there is no
difference, we conclude that the estimated difference of 3 percent is
not statistically significant. 


   CONTROLLING FOR NONSAMPLING
   ERRORS
--------------------------------------------------------- Appendix I:4

In addition to the reported sampling errors, various obstacles can
occur when conducting this type of review and may cause other types
of errors, commonly referred to as nonsampling errors.  For example,
differences in how questions are interpreted and errors in entering
data could affect the results.  We included steps in both the data
collection and data analysis stages for the purpose of minimizing
such nonsampling errors.  These steps involved the 100 percent review
of completed data collection instruments (DCI) and data entry of
those DCIs, and checking all computer analyses with a second analyst. 

Tables I.3 through I.5 describe our point estimates for the analysis
of financial status audits and the related sampling errors. 



                                    Table I.3
                     
                     Confidence Intervals for Point Estimates
                        Comparing 1992-1993 and 1995-1996
                             Financial Status Audits

                      1992-1993           1995-1996            Comparison\a
                  ------------------  ------------------  ----------------------
                          Confidence          Confidence              Confidence
                            interval            interval                interval
                          at the 95-          at the 95-              at the 95-
                             percent             percent  Percentage     percent
                  Estima  confidence  Estima  confidence  Difference  confidence
Description           te       level      te       level          \c       level
----------------  ------  ----------  ------  ----------  ----------  ----------
Audits where a       24%        + 5%     22%        + 4%          3%        + 6%
 financial
 status
 technique(s)
 was used
 (percentage of
 all audits)
Audits using         75%        +10%     84%        +10%          9%        +14%
 financial
 status
 techniques that
 had business or
 farm income (as
 a percentage of
 financial
 status audits)
Percent of           38%         +7%     34%         +7%          4%        +10%
 returns with
 business or
 farm income
 where one or
 more of the
 financial
 status
 techniques were
 used
Percent of           12%         +6%      7%         +5%          4%         +8%
 returns with
 nonbusiness
 income where
 one or more of
 the financial
 status
 techniques were
 used

Initial interview questions the AICPA considers inappropriate (as a percent of
all audits with initial interviews documented)
--------------------------------------------------------------------------------
Taxpayer's            7%        + 3%     12%        + 5%          5%        + 6%
 education
Assets other          6%        + 3%      8%        + 4%          2%        + 5%
 than home or
 autos that cost
 over $10,000
Loans by and          8%        + 3%      4%        + 2%          3%        + 4%
 loan payments
 to the taxpayer
Amount and            9%        + 5%      6%        + 3%          3%        + 6%
 monthly
 payments on
 outstanding
 debt
Cash advances         3%        + 2%      1%        + 1%          2%        + 2%
 from credit
 cards
Amount of cash       27%         +6%     24%        + 6%          3%         +9%
 on hand
Amounts               \b          \b      1%        + 1%          \b        + 1%
 transferred
 between
 accounts
Safe deposit box     20%        + 6%     18%        + 6%          2%        + 8%
Taxpayer              4%        + 3%      3%        + 2%          1%        + 3%
 involved in
 transactions of
 $10,000 or more
Information           \b          \b      3%        + 2%          3%        + 2%
 about where the
 taxpayer
 vacations
Information           1%        + 1%      1%        + 1%          \b        + 2%
 about what
 college the
 taxpayer's
 children attend
Quality of the        0%          0%      0%          0%          0%          0%
 taxpayer's
 clothing
Information           1%        + 2%      1%        + 1%          \b        + 2%
 about how often
 the taxpayer
 eats out
Information           \b          \b      1%        + 1%          1%        + 1%
 about how much
 the taxpayer
 spends on
 entertainment
Information on        7%        + 4%     11%        + 5%          4%        + 6%
 taxpayer's cash
 horde
Information on        3%        + 3%      1%        + 1%          1%        + 3%
 the amount the
 taxpayer paid
 for utilities
 and personal
 living expenses
--------------------------------------------------------------------------------
\a For the comparisons between years, when the confidence interval of
the difference is greater than the difference, the result is not
statistically significant. 

\b Percentages are less than 0.5 percent. 

\c Percent of difference may not add to total due to rounding. 

Source:  GAO analysis of sampled data. 



                                    Table I.4
                     
                     Comparison of Adjustments to Income For
                     1995 Between Audits That Used Financial
                     Status Techniques Versus Audits That Did
                                       Not

        Financial status        No financial status           Comparison\a
    ------------------------  ------------------------  -------------------------
                                                                    Confidence
               Confidence                Confidence               interval at the
            interval at the           interval at the               95-percent
               95-percent                95-percent                 confidence
            confidence level          confidence level                 level
            ----------------          ----------------            ---------------
De
sc
ri
pt              Low     High              Low     High               Low     High
io  Estima  estimat  estimat  Estima  estimat  estimat  Differen  estima  estimat
n       te        e        e      te        e        e        ce      te        e
--  ------  -------  -------  ------  -------  -------  --------  ------  -------
1995 sample results
---------------------------------------------------------------------------------
Ad  $14,73  $11,041  $19,622  $8,348   $6,998   $9,974    $6,384  $3,162  $12,882
 j       2
 u
 s
 t
 m
 e
 n
 t
 s
 t
 o
 i
 n
 c
 o
 m
 e
---------------------------------------------------------------------------------
\a For the comparison between audits that used financial status
techniques and those that did not, when the confidence interval of
the difference is greater than the difference, the result is not
statistically significant. 

Source:  GAO analysis of sampled data. 



                                    Table I.5
                     
                      Confidence Intervals for Estimates of
                     Adjustments to Income Attributed to the
                        Use of Financial Status Techniques

                              (Dollars in millions)

                   1992-1993 audits                    1995-1996 audits
          ----------------------------------  ----------------------------------
                      Confidence interval at              Confidence interval at
                          the 95-percent                      the 95-percent
                         confidence level                    confidence level
                      ----------------------              ----------------------
Descript
ion of     Estimated         Low        High   Estimated         Low        High
variable     dollars    estimate    estimate     dollars    estimate    estimate
--------  ----------  ----------  ----------  ----------  ----------  ----------
Total         $1,154        $382      $3,483       $316\        $188        $533
 adjustm
 ents to
 income
 identif
 ied by
 financi
 al
 status
 techniq
 ues
--------------------------------------------------------------------------------
Source:  GAO analysis of sampled data. 



                                    Table I.6
                     
                       Confidence Intervals for Estimate of
                       Variables Without Comparisons--1992-
                            1993 and 1995-1996 Audits

                            1992-1993 audits              1995-1996 Audits
                      ----------------------------  ----------------------------
                                        Confidence                    Confidence
                                   interval at the               interval at the
Description of        Percentage        95-percent  Percentage        95-percent
Variables               Estimate  confidence level    Estimate  confidence level
--------------------  ----------  ----------------  ----------  ----------------
Audits using                 23%               +9%         28%              +10%
 financial status
 techniques that
 were closed with no
 change to tax or
 income
Audits using                 81%              + 8%         83%              + 6%
 financial status
 techniques that
 were closed with no
 change identified
 by one of the
 techniques
Adjustments of less          12%              + 7%          8%              + 5%
 than $10,000
 identified by
 financial status
 techniques
Adjustments of                7%              + 4%          9%              + 4%
 $10,000 or more
 identified by
 financial status
 techniques
Audits where                  9%              + 3%          6%              + 3%
 supervisory review
 was documented in
 the workpapers
Audits with                   0%              + 0%          7%              + 7%
 supervisory review
 where the
 supervisor
 mentioned financial
 status
Time examiners                2%              + 3%          3%              + 2%
 requested personal
 living expenses
 from taxpayers in a
 notification letter
Preliminary Cash-T           n/a               n/a         23%              +10%
 was the only
 financial status
 technique
Technique used was           50%              +10%         47%              +11%
 Cash-T only
Technique used was           29%              + 8%         21%              + 9%
 bank deposit only
Technique used was           20%               +8%         32%              + 9%
 combination of Cash
 T and bank deposit
Other technique was           \a                \a          1%              + 1%
 used
--------------------------------------------------------------------------------
\a Percentages are less than 0.5 percent. 

Source:  GAO analysis of sampled data. 


IRS' EXAMINATION QUALITY
MEASUREMENT SYSTEM
========================================================== Appendix II

The Office of Compliance Specialization, within IRS' Examination
Division, has responsibility for Quality Measurement Staff operations
and the Examination Quality Measurement System (EQMS).  Among other
uses, IRS uses EQMS to measure the quality of closed audits against
nine IRS audit standards.  The standards address the scope, audit
techniques, technical conclusions, workpaper preparation, reports,
and time management of an audit.  Each standard includes additional
key elements describing specific components of a quality audit. 
Table II.1 summarizes the standards and the associated key elements. 



                                    Table II.1
                     
                       Summary of IRS' Examination Quality
                      Measurement System Auditing Standards
                               (as of October 1996)

No.   Standard          Key Elements      Purpose           Overview
----  ----------------  ----------------  ----------------  --------------------
1     Considered        A. Balance sheet  Measures whether  This standard
      large, unusual,   and Schedule M    consideration     encompasses, but is
      or questionable   considered        was given to the  not limited to, the
      items             B. Income,        large, unusual,   following
                        deduction, and    or questionable   fundamental
                        credit items      items in both     considerations:
                        considered        the precontact    absolute dollar
                        C. Scope of       stage and during  value, relative
                        examination was   the course of     dollar value,
                        appropriate       the examination.  multiyear
                                                            comparisons, intent
                                                            to mislead,
                                                            industry/business
                                                            practices,
                                                            compliance impact,
                                                            and so forth.

2     Probes for        A. Consideration  Measures whether  Gross receipts were
      unreported        of internal       the steps taken   probed during the
      income            controls for all  verified that     course of
                        business          the proper        examination,
                        returns           amount of income  regardless of
                        B. Consideration  was reported.     whether the taxpayer
                        of books and                        maintained a double
                        records                             entry set of books.
                        C. Consideration                    Consideration was
                        of financial                        given to responses
                        status                              to interview
                        D. Appropriate                      questions, the
                        use of indirect                     financial status
                        methods                             analysis, tax return
                                                            information, and the
                                                            books and records in
                                                            probing for
                                                            unreported income.

3     Required filing   A. Consideration  Measures whether  Required filing
      checks            of prior and      consideration     checks consist of
                        subsequent year   was given to      the analysis of
                        tax returns       filing and        return information
                        B. Consideration  examination       and, when warranted,
                        of related        potential of all  the pick-up of
                        returns           returns required  related, prior and
                        C. Compliance     by the taxpayer   subsequent year
                        items considered  including those   returns. In
                                          entities in       accordance with
                                          taxpayer's        Internal Revenue
                                          sphere of         Manual 4034,
                                          influence/        examinations should
                                          responsibility.   include checks for
                                                            filing information
                                                            returns.

4     Examination       A. Adequate       Measures whether  The depth of the
      depth and         interviews        the issues        examination was
      records examined  conducted         examined were     determined through
                        B. Adequate exam  completed to the  inspection, inquiry,
                        techniques used   extent necessary  interviews,
                        C. Fraud          to provide        observation, and
                        adequately        sufficient        analysis of
                        considered and    information to    appropriate
                        developed         determine         documents, ledgers,
                        D. Issues         substantially     journals, oral
                        sufficiently      correct tax.      testimony, third-
                        developed                           party records, etc.,
                                                            to ensure full
                                                            development of
                                                            relevant facts
                                                            concerning the
                                                            issues of merit.
                                                            Interviews provided
                                                            information not
                                                            available from
                                                            documents to obtain
                                                            an understanding of
                                                            the taxpayer's
                                                            financial history,
                                                            business operations,
                                                            and accounting
                                                            records in order to
                                                            evaluate the
                                                            accuracy of books/
                                                            records. Specialists
                                                            provided expertise
                                                            to ensure proper
                                                            development of
                                                            unique or complex
                                                            issues.

5     Findings          A. Correct        Measures whether  This standard
      supported by law  technical/        the conclusions   includes
                        factual           reached were      consideration of
                        conclusions       based on a        applicable law,
                        reached           correct           regulations, court
                                          application of    cases, revenue
                                          tax law.          rulings, etc. to
                                                            support technical/
                                                            factual conclusions.

6     Penalties         A. Recognized,    Measures whether  Consideration of the
      properly          considered, and   applicable        application of
      considered        applied           penalties were    appropriate
                        correctly         considered and    penalties during all
                        B. Penalties      applied           examination is
                        computed          correctly.        required.
                        correctly

7     Workpapers        A. Fully          Measures the      Workpapers provided
      support           disclose audit    documentation of  the principal
      conclusions       trail and         the               support for the
                        techniques        examination's     examiner's report
                        B. Legible and    audit trail and   and documented the
                        organized         techniques used.  procedures applied,
                        C. Adjustments                      tests performed,
                        in workpapers                       information
                        agree with 4318,                    obtained, and the
                        4700, and                           conclusions reached
                        reports                             in the examination.
                        D. Activity
                        record
                        adequately
                        documents exam
                        activities
                        E. Disclosure

8     Report writing    A. Applicable     Measures the      Addresses the
      procedures        report writing    presentation of   written presentation
      followed          procedures        the audit         of audit findings in
                        followed          findings in       terms of content,
                        B. Correct tax    terms of          format, and
                        computation       content, format,  accuracy. All
                                          and accuracy.     necessary
                                                            information is
                                                            contained in the
                                                            report, so that
                                                            there is a clear
                                                            understanding of the
                                                            adjustments made and
                                                            the reasons for
                                                            those adjustments.

9     Time span/time    A. Examination    Measures the      Time is an essential
      charged           time              utilization of    element of the
                        commensurate      time as it        Auditing Standards
                        B. Exam           relates to the    and is a proper
                        initiation        complete audit    consideration in
                        C. Examination    process.          analyses of the
                        activities                          examination process.
                        D. Case closing                     The process is
                                                            considered as a
                                                            whole and at
                                                            examination
                                                            initiation,
                                                            examination
                                                            activities, and case
                                                            closing stages.
--------------------------------------------------------------------------------
Source:  IRS data. 


      STANDARD SUCCESS RATE
------------------------------------------------------ Appendix II:0.1

EQMS quality reviewers use the key element definitions to determine
whether an audit adhered to the standard.  Thus, adherence to audit
quality is measured by the presence or absence of associated key
elements.  For a standard to be rated as having been met, each of the
associated key elements must also be rated as met or not applicable. 
If the audit does not demonstrate the characteristics described by
one of the key elements, then the standard is rated as not met. 

One measure that IRS uses to evaluate the audit quality is the
standard success rate.  It measures the percentage of cases for which
all the underlying key elements of each standard are rated as having
been met.  According to IRS, this measure is useful for determining
whether a case is flawed and in what area.  Figures II.1 and II.2
show the standard success rates for each of the standards for fiscal
years 1992 through 1996 for office and field audits, respectively. 

   Figure II.1:  Standard Success
   Rates for Office Audits From
   Fiscal Years 1992-1996

   (See figure in printed
   edition.)

Source:  IRS data. 

   Figure II.2:  Standard Success
   Rates for Field Audits From
   Fiscal Years 1992-1996

   (See figure in printed
   edition.)

Source:  IRS data. 


      KEY ELEMENT PASS RATE
------------------------------------------------------ Appendix II:0.2

IRS also uses the key element pass rate as a measure of audit
quality.  This measure computes the percentage of audits
demonstrating the characteristics defined by the key element. 
According to IRS, the key element pass rate is the most sensitive
measurement and is useful when describing how an audit is flawed,
establishing a baseline for improvement, and identifying systemic
changes.  Figures II.3 and II.4 show the pass rates for the key
elements of Standard 2 for fiscal years 1992 through 1996 for office
and field audits, respectively. 

   Figure II.3:  Key Element Pass
   Rates for Key Elements of
   Standard 2 for Office Audits
   From Fiscal Years 1992-1996

   (See figure in printed
   edition.)

Source:  IRS data. 

   Figure II.4:  Key Element Pass
   Rates for Key Elements of
   Standard 2 for Field Audits
   From Fiscal Years 1992-1996

   (See figure in printed
   edition.)

Source:  IRS data. 


ANALYSIS OF AICPA CONCERNS
========================================================= Appendix III

The American Institute of Certified Public Accountants (AICPA) has
been among the critics of IRS' reemphasis on financial status audits
since the program began in late 1994.  During 1995 and 1996,
officials from IRS and AICPA met several times to discuss these
concerns and, to some extent, IRS mitigated the problems with memos
clarifying the use of financial status techniques.  AICPA has had a
long list of concerns about actions taken by IRS auditors, including

  -- sending a personal living expense (PLE) form with the letter
     notifying taxpayers of the audit before finding any evidence of
     underreported income;

  -- asking financial status questions at the initial interview,
     before having any evidence of underreported income;

  -- arriving unannounced to inspect a personal residence;

  -- bypassing a valid power of attorney and requesting information
     or records directly from taxpayers;

  -- interviewing taxpayers without the presence of their
     representative; and

  -- requiring taxpayers' representative to submit a freedom of
     information request to obtain third-party documents on their
     clients. 

Neither AICPA or IRS had any objective data on these concerns.  Using
our sample, however, we were able to collect data on the first two
concerns involving PLE forms and financial status questions. 

As for the PLE forms, AICPA indicated that some audit notification
letters asked taxpayers to complete this form even though IRS had no
evidence of underreported income.  AICPA officials believed this
request was intrusive, burdensome, and costly to taxpayers.  The
officials said PLE information should be requested only after IRS had
some objective evidence that taxpayers had underreported income on
tax returns. 

In reviewing the workpapers for our two samples, we looked for copies
of notification letters.  We found very few examples in which the
letters asked taxpayers to complete a PLE form.  On the basis of our
sample, we estimate that IRS used the notification letter to request
PLE forms in no more than 5 percent of the audits for both the 1992
and 1993 and the 1995 and 1996 samples. 

In a March 1996 memorandum, the Acting Assistant Commissioner
(Examination) clarified the PLE instructions.  The memorandum
indicated that while auditors had the responsibility to secure an
overall financial picture of the taxpayer, they were not expected to
automatically request PLE information with the notification letter. 
According to AICPA officials, sending PLE forms with the notification
letters has decreased since the distribution of this memorandum. 

AICPA officials were also concerned that auditors were asking
personal questions about the taxpayer's financial status at the
initial interview before having any evidence of underreported income. 
Auditors use the initial interview to explain the audit process, the
taxpayer's rights, and gain an understanding of the taxpayer's
situation.  Generally, auditors prepare workpapers to summarize these
interviews.  We reviewed these interview write-ups and collected data
on the types of questions asked by auditors at this meeting. 

AICPA officials identified questions that caused them concern.  We
collected information on whether the auditors asked these questions
both before and after IRS began reemphasizing financial status.  We
compared these two periods because AICPA had associated the questions
with the renewed emphasis by IRS on financial status audits, and the
1992 and 1993 period was just prior to this renewed emphasis.  Table
III.1 shows how often auditors asked these questions at initial
interviews in 1992 and 1993 and in 1995 and 1996 audits. 



                              Table III.1
                
                 Questions Asked by Auditors at Initial
                  Interviews, Which Cause Concerns for
                            AICPA Officials

                                            Percent of time question
                                                     asked
                                          ----------------------------
Initial interview questions                   1992-1993      1995-1996
----------------------------------------  -------------  -------------
Educational background of taxpayers                   6             12
Assets other than home or autos, that                 6              8
 cost over $10,000
Loans by and loan payments to the                     8              4
 taxpayer
Amount and monthly payments on                        9              6
 outstanding debt
Cash advances from credit cards                       3              1
Amount of cash on hand                               27             24
Amounts transferred between accounts                 <1              1
Safe deposit box                                     20             18
Taxpayer involved in transactions of                  4              3
 $10,000 or more
Information about where the taxpayer                 <1              3
 vacations
Information about what college the                    1              1
 taxpayer's children attend
Information about the quality of the                  0              0
 taxpayer's clothing
Information about how often the taxpayer              1              1
 eats out
Information about how much the taxpayer              <1              1
 spends on entertainment
Information on the taxpayer's cash horde              7             11
Information on the amount the taxpayer                3              1
 paid for utilities and personal living
 expenses
----------------------------------------------------------------------
Source:  Analysis of GAO samples of IRS audits for 1992 and 1993 and
for 1995 and 1996. 

As shown in table III.1, with few exceptions, little difference
exists in how often these questions were asked at initial interviews
in 1992 and 1993 and in 1995 and 1996 audits. 

In his March 1996 memorandum to Regional Chief Compliance Officers,
the Acting Assistant Commissioner (Examination) provided general
guidance on how far to probe for unreported income at the initial
interview.  He emphasized that auditors must evaluate the facts and
use judgment.  The memo further stated that performing in-depth
income probes and asking questions about personal assets and
expenditures were not effective uses of resources without a
reasonable indication of unreported income. 




(See figure in printed edition.)Appendix IV
COMMENTS FROM THE INTERNAL REVENUE
SERVICE
========================================================= Appendix III



(See figure in printed edition.)



(See figure in printed edition.)


The following are GAO's comments on the Internal Revenue Service's
letter dated November 20, 1997. 

GAO COMMENTS

1.  IRS suggested that we change the title of the report to respond
to the first objective of our work and suggested a title that would
point out that IRS has not increased the use of financial status
techniques.  IRS believed that by focusing on the need for more
criteria, readers of the report would infer that IRS was being
unnecessarily intrusive.  We considered changing the title but
decided against it for various reasons.  First, our report already
discussed the issue of intrusiveness, pointing out that use of the
techniques did not necessarily mean intrusions into taxpayers'
affairs, particularly when such usage identified changes to reported
income.  Second, such a title would ignore the other three objectives
of our report.  We concluded that the focus on the need for more
criteria not only could be associated with all four objectives but
also with the actions needed to prompt improvements. 

2.  IRS said that the report cited no evidence of any increased
intrusiveness and that the fact that use of the techniques led to no
tax change does not diminish Examination's responsibility to
determine the correct tax liability.  We believe that IRS
misinterpreted our discussion of intrusiveness.  In the draft report,
we noted that the reason for no evidence of intrusiveness was that it
was not available from IRS or others.  We observed, however, that
only the preliminary Cash-T results in no additional burden on the
taxpayer, while the burden imposed through the use of other
techniques varies depending on the amount of additional taxpayer
contact.  Also, our draft did not say that there is any relationship
between the no change rate and IRS' responsibility to determine the
correct tax liability.  Accordingly, we made no changes to the report
to reflect these comments. 


MAJOR CONTRIBUTORS TO THIS REPORT
=========================================================== Appendix V

GENERAL GOVERNMENT DIVISION,
WASHINGTON, D.C. 

Thomas D.  Short, Assistant Director, Tax Policy and Administration
Issues
Patricia H.  McGuire, Assistant Director
Tim Outlaw, Senior Evaluator
James M.  Fields, Technical Advisor
Susan F.  Baker, Computer Specialist

SAN FRANCISCO REGIONAL OFFICE

Louis G.  Roberts, Evaluator-in-Charge
Kathleen E.  Seymour, Senior Evaluator
Samuel H.  Scrutchins, Senior Data Analyst

*** End of document. ***