Tax Administration: IRS' 1998 Tax Filing Season (Letter Report, 12/31/98,
GAO/GGD-99-21).

Internal Revenue Service (IRS) indicators show that IRS generally met or
exceeded its performance goals for the 1998 tax filing season. Included
in IRS indicators are measures of the timeliness of its processing of
refunds, the accessibility of its telephone service, the accuracy with
which it processes returns, and the accuracy of the assistance it gives
over the telephone. Although millions of taxpayers used the services
provided by IRS' walk-in sites during the 1998 tax filing season, the
agency lacked meaningful nationwide data for assessing the performance
of those sites. This report also discusses six specific areas: (1) IRS
efforts to increase the use of electronic filing; (2) IRS' progress in
addressing an issue discussed in GAO reports on the 1996 and 1997 filing
seasons involving the use of private banks to process some tax payments;
(3) IRS' implementation of the 1997 tax law change dealing with capital
gains; (4) the status of IRS' efforts to reduce Earned Income Credit
noncompliance; (5) the ability of taxpayers seeking assistance to reach
IRS by telephone; and (6) other IRS efforts to provide information to
taxpayers, such as its web site on the Internet and TeleTax. GAO also
discusses IRS' test of a new processing system that has implications for
future filing seasons.

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

 REPORTNUM:  GGD-99-21
     TITLE:  Tax Administration: IRS' 1998 Tax Filing Season
      DATE:  12/31/98
   SUBJECT:  Tax administration
             Tax returns
             Electronic forms
             Performance measures
             Tax refunds
             Tax violations
             Customer service
             Telephone
IDENTIFIER:  IRS TeleFile Program
             IRS Electronic Filing System
             IRS 1040PC Program
             Earned Income Tax Credit
             IRS TeleTax System
             EIC
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  This text was extracted from a PDF file.        **
** Delineations within the text indicating chapter titles,      **
** headings, and bullets have not been preserved, and in some   **
** cases heading text has been incorrectly merged into          **
** body text in the adjacent column.  Graphic images have       **
** not been reproduced, but figure captions are included.       **
** Tables are included, but column deliniations have not been   **
** preserved.                                                   **
**                                                              **
** 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.                         **
******************************************************************
gg99021 TAX ADMINISTRATION

IRS' 1998 Tax Filing Season

United States General Accounting Office

GAO Report to the Chairman, Subcommittee on Oversight, Committee
on Ways and

Means, House of Representatives


December 1998 

GAO/GGD-99-21

December 1998   GAO/GGD-99-21

United States General Accounting Office Washington, D. C. 20548

General Government Division

B-279310

Page 1 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

GAO December 31, 1998 The Honorable Nancy L. Johnson Chairman,
Subcommittee on Oversight Committee on Ways and Means House of
Representatives

Dear Chairman Johnson: This report responds to your request that
we assess the Internal Revenue Service's (IRS) performance during
the 1998 tax filing season. In addition to providing data on
various indicators that IRS uses to measure its filing season
performance, we discuss six areas about which you specifically
inquired: (1) IRS' efforts to increase the use of electronic
filing; (2) IRS' progress in addressing an issue discussed in our
reports on the 1996 and 1997 filing seasons involving the use of
private banks to process some tax payments; 1 (3) IRS'
implementation of the 1997 tax law change dealing with capital
gains; (4) the status of IRS' efforts to reduce Earned Income
Credit (EIC) noncompliance; (5) the ability of taxpayers seeking
assistance to reach IRS by telephone; and (6) other IRS efforts to
provide information to taxpayers, such as its World Wide Web site
on the Internet and TeleTax (an automated system that provides
recorded information on a variety of tax topics). Finally, we
discuss IRS' test of a new processing system that has implications
for future filing seasons. In March 1998, we testified before the
Subcommittee on the interim results of our work. 2

IRS has several indicators that it uses to judge the success of a
filing season. For the 1998 filing season, those indicators showed
that IRS generally met or exceeded its performance goals. Included
in IRS' indicators are measures of the timeliness of its
processing of refunds, the accessibility of its telephone service,
the accuracy with which it processes returns, and the accuracy of
assistance it provides over the telephone. Although millions of
taxpayers used the services provided by IRS' walk- in sites during
the 1998 filing season, IRS did not have meaningful nationwide
data for assessing the performance of those sites.

1 IRS' 1996 Tax Filing Season: Performance Goals Generally Met;
Efforts to Modernize Had Mixed Results (GAO/GGD-97-25, Dec. 18,
1996) and Tax Administration: IRS' 1997 Tax Filing Season
(GAO/GGD-98-33, Dec. 29, 1997).

2 Tax Administration: IRS' Fiscal Year 1999 Budget Request and
Fiscal Year 1998 Filing Season (GAO/ TGGD/ AIMD- 98- 114, Mar. 31,
1998). Results in Brief

B-279310 Page 2 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

This report addresses several other matters related to the 1998
filing season. For example:

 Although the overall number of individual income tax returns
filed in 1998 increased by less than 2 percent from 1997, the
number of returns filed electronically, including the number filed
over the telephone (i. e., TeleFile), increased at a more robust
rate about 28 percent. According to IRS and our analysis, the
increase in electronic filing was due, in part, to (1) a decision
by the largest nationwide return preparation company to include
electronic filing in its basic return preparation fee and (2) a
change in IRS' procedures that made more persons eligible to use
TeleFile in 1998.

 Even with the 28- percent increase in electronic filing, about
98.5 million returns (80 percent) were filed on paper in 1998.
IRS' most recent survey of taxpayers who were eligible to use
TeleFile, but who chose not to, identified one possible step IRS
could take to increase the use of that filing alternative. Over 70
percent of the respondents to that survey said that they would
have been encouraged to file by telephone if the special tax
package they were sent included the tax table. IRS has opposed
including the tax table in the package because doing so would
increase the size and cost of the package and because taxpayers do
not need the table to file by telephone. Although IRS is correct
on both counts, it has not conducted a test to determine whether
the potential increase in TeleFile use would justify the
additional cost of including the tax table in the package.

 IRS uses private banks, known as lockboxes, to process some tax
payments submitted with Forms 1040 (U. S. Individual Income Tax
Return). As part of that arrangement, IRS requires that taxpayers
send their tax returns, along with their payments, to the banks.
The government pays the banks several million dollars to sort
those returns and ship them to IRS for processing. As we reported
in 1997, IRS has inadequate evidence to justify the additional
cost associated with having banks handle tax returns.

 The return processing procedures that IRS adopted in implementing
a legislative change relating to capital gains led to processing
delays and increased taxpayer burden in 1998. IRS plans to revise
its procedures for the 1999 filing season. The revisions, if
effectively implemented, should alleviate some of the problems
encountered in 1998.

 IRS expanded its efforts to ensure that taxpayers filed correct
EIC claims in 1998. However, EIC criteria may not have been
applied consistently at service centers due to a lack of controls
and standardized procedures to ensure that field offices receive
timely notification of procedural changes. Also, as we previously
reported, IRS could have done more to alert taxpayers to the
consequences of falsely claiming the EIC and could have

B-279310 Page 3 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

provided special EIC assistance earlier in the filing season. 3
IRS plans changes in 1999 that partially address these issues, but
more can be done.

 The measures that IRS uses to judge the accessibility of its
telephone service showed substantial improvement. For example, the
level of service, which identifies the percentage of calls that
received an answer, increased from 52.6 percent in 1997 to 73.7
percent in 1998. The results of our test of access to IRS' tax
assistance telephone number were consistent with IRS' reported
increase. IRS took several steps in 1998 that may have contributed
to the improved accessibility, including (1) extending the hours
during which taxpayers could call IRS and speak to an assistor,
(2) expanding its use of voice messaging, and (3) increasing the
number of staff available to answer the telephone during peak
calling periods.

 Use of IRS' Web site, as measured by the number of hits,
increased by 187 percent this year, while the number of files
downloaded from that site increased by about 313 percent. We
tested the downloading of forms and publications during the year
and consistently found that the process was quick and easy.

 Use of TeleTax decreased about 15 percent during the 1998 filing
season. Customer satisfaction surveys show that a majority of the
users of TeleTax expressed some dissatisfaction with that service.
However, the survey only asked if users were satisfied or
dissatisfied; it did not ask any follow- up questions that could
be used to determine why a respondent might have been
dissatisfied.

We also monitored IRS' test of a new system for processing returns
and remittances. Because of problems with the part of the system
that processes remittances, IRS has decided to revise the system's
implementation schedule and its contingency plan for 1999. We are
concerned about the potential impact of the revised schedule on
the 2000 filing season and the absence of a contingency plan for
2000, when current systems will no longer work.

Our objective was to assess IRS' performance during the 1998
filing season, with particular emphasis on several areas
identified in the Subcommittee's request. To achieve our
objective, we

 interviewed IRS officials at the National Office; the Midstates,
Southeast, and Western Regional Offices; the Georgia, Kansas/
Missouri, and Northern

3 Earned Income Credit: IRS' Tax Year 1994 Compliance Study and
Recent Efforts to Reduce Noncompliance (GAO/GGD-98-150, July 28,
1998). Objective, Scope, and

Methodology

B-279310 Page 4 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

California District Offices; and the Atlanta, GA; Austin, TX;
Cincinnati, OH; Fresno, CA; and Kansas City, KS, Service Centers;
4

 interviewed officials from the two largest national tax return
preparation companies;

 analyzed filing- season- related data from various management
information systems, such as IRS' Management Information System
for Top Level Executives; IRS data on processing errors, including
errors involving the EIC and capital gains; and IRS data on
alternative filing methods and TeleTax;

 interviewed IRS National Office and service center officials who
were responsible for implementing a new system for processing
returns and remittances and analyzed reports on a test of that
system at the Austin Service Center;

 analyzed IRS data relating to its toll- free telephone assistance
and conducted a nonstatistical test of IRS' telephone
accessibility during the filing season (see app. I for information
on our test methodology and detailed results);

 interviewed officials who were responsible for walk- in
assistance sites in IRS' Southeast and Midstates Regional Offices
and officials at walk- in sites in Atlanta; Mission, KS; and St.
Louis, MO; 5

 reviewed walk- in site guidance developed by IRS' National
Office; the results of walk- in site reviews done by one regional
office; and a contractor's report on the results of surveys of
walk- in customers done in March, April, and May, 1998;

 analyzed activity data for IRS' forms distribution centers and
World Wide Web site and conducted a nonstatistical test of the
ability to download files from the Web site;

 interviewed officials from IRS and the Department of the
Treasury's Financial Management Service (FMS) (which is
responsible for negotiating and administering lockbox contracts)
about the use of lockboxes to process Form 1040 tax payments and
analyzed cost/ benefit data related to lockbox processing;

 interviewed officials from IRS' Taxpayer Advocate's Office about
the impact of various filing season activities on taxpayers; and

 reviewed reports by IRS' Office of Internal Audit on filing
season activities and a report by Treasury's Inspector General on
the use of lockboxes.

4 Except for Austin, we selected these locations because we had
staff available in those areas to do the work. We selected the
Austin Service Center because it was responsible for testing IRS'
new processing system.

5 We selected these locations because we had staff available to do
the work in those cities.

B-279310 Page 5 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

We did our work from January through October, 1998, in accordance
with generally accepted government auditing standards.

Table 1 shows various indicators that IRS used to measure its
performance during the 1998 filing season. Those indicators, which
we selected from various IRS documents and sources, relate to
timeliness, such as the number of days needed to process and issue
refunds; quality, such as the accuracy of answers to taxpayers'
questions and notices sent to taxpayers; and service
accessibility, such as the extent to which taxpayers with
taxrelated questions were able to reach IRS by telephone. 6 The
goals shown in table 1 were set by IRS generally on the basis of
historical accomplishments and projected workload. We did not
assess the appropriateness of IRS' goals.

Table 1 shows 11 indicators. However, IRS added one indicator, the
level of service provided by the taxpayer service telephone
system, after planning for the 1998 filing season had been
completed and thus did not establish a goal for that indicator. Of
the 10 indicators for which IRS had established goals for 1998,
IRS met or exceeded the goals for 7 indicators and came close to
the goals for the other 3 indicators. Table 1 also shows that for
8 of the 11 indicators, IRS exceeded its performance in 1997. IRS
equaled its 1997 performance for one indicator and saw its
performance on the other two indicators drop slightly.

6 IRS also has various workload indicators, such as the number of
returns received and refunds issued. Several of those indicators,
which generally show a growth in IRS' filing season workload and a
growth in taxpayer use of fairly new alternatives, such as
electronic filing, direct deposits, and IRS' World Wide Web site,
are shown in appendix II. IRS Generally Met the

Filing Season Goals It Established

B-279310 Page 6 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

1997 a 1998 a Indicator Goal Accomplishment Goal Accomplishment

Accuracy of individual income tax returns processed by Code and
Edit staff b

Process 95% accurately 95% were processed accurately Process 95%
accurately 95.5% were processed

accurately Accuracy of individual income tax returns processed by
data transcribers c

Process 95% accurately 94.7% were processed accurately Process 95%
accurately 94.2% were processed

accurately Notice accuracy d Provide accurate

notices to taxpayers 98.5% of the time

Provided accurate notices to taxpayers 98.3% of the time

Provide accurate notices to taxpayers 98.5% of the time

Provided accurate notices to taxpayers 98.4% of the time
Timeliness of processing tax payments submitted with individual
income tax returns e

Payments received 4/ 15/ 97 through 4/ 30/ 97 were to be deposited
no later than 4/ 30/ 97

All payments received 4/ 15/ 97 through 4/ 30/ 97 were deposited
by 5/ 2/ 97

Payments received 4/ 15/ 98 through 4/ 30/ 98 were to be deposited
no later than 4/ 30/ 98

All payments received 4/ 15/ 98 through 4/ 30/ 98 were deposited
by 4/ 30/ 98 Accuracy of individual income tax refunds on paper
returns f Process 99.3%

accurately 99.4% were processed accurately Process 99.3%

accurately 99.6% were processed accurately Timeliness of refund
checks for individual income tax returns filed on paper g

Issue within an average of 40 days Issued within an

average of 38 days as of 5/ 97

Issue within an average of 40 days Issued within an

average of 35 days as of 5/ 98 Timeliness of refunds for
individual income tax returns filed electronically h

Issue within an average of 21 days Issued within an

average of 15 days as of 5/ 97

Issue within an average of 21 days Issued within an

average of 15 days Level of service provided by taxpayer service
telephone system i

Not applicable j Provided 52. 6% level of service Not applicable j
Provided 73. 7% level of

service Level of access to taxpayer service telephone system k
Provide 60. 2% level of

access Provided 71% level of access k Provide 70% level of

access Provided 90. 6% level of access Accuracy of tax law
assistance l Answer 92% accurately 95% were answered

accurately Answer 96% accurately 93.6% were answered accurately
Accuracy of processing form orders m Process 96.5%

accurately 97% were processed accurately Process 96.5%

accurately 97.3% were processed accurately Note: Similar tables in
our reports on past filing seasons included three indicators that
are not included in this year's table. We dropped the indicator
service center individual income tax returns processing
productivity because IRS stopped reporting that measure. We also
dropped the indicator individual income tax returns processing
cycle time because we determined that the refund timeliness
measures provide more meaningful information on the timeliness of
IRS processing. Lastly, the indicator level of access to forms-
ordering telephone system was dropped because IRS now includes
telephone calls to its forms distribution centers in computing the
taxpayer service level of access and level of service indicators.
a Data are as of April 1997 and April 1998, unless otherwise
noted.

b Code and Edit staff are to prepare returns for computer entry
by, among other things, ensuring that all data are present on the
return and legible. This indicator represents the percentage of
other than full paid, individual income tax returns processed
without code and edit errors. Other than full paid returns are
those that involve either a refund or an unpaid liability. c This
indicator represents the percentage of other than full paid,
individual income tax returns

processed without transcription errors. d This indicator is based
on a sample of returns processing notices addressed to individual
and

business taxpayers. Among other things, returns processing notices
are used to notify taxpayers of missing schedules or forms,
missing Social Security numbers, or refunds being delayed or used
to offset another liability. IRS compares the printed notice to
various data, including information in the taxpayer's master file
record and on the taxpayer's tax return. The indicator is
calculated by dividing

Table 1: IRS' Performance Goals and Related Accomplishments for
the 1997 and 1998 Filing Seasons

B-279310 Page 7 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

the number of correct notices reviewed by the total number of
notices reviewed. IRS told us that the results for individual and
business taxpayers cannot be separated. e Service centers are to
deposit payments in a timely manner, generally within 24 hours of
receipt.

Because of the volume of payments received between April 15 and
April 30, IRS suspends the 24- hour requirement during that
period. Instead, IRS requires that all payments received during
that period are to be deposited by the end of business on April
30. After April 30, the centers are to resume the 24- hour deposit
schedule. f This indicator is based on a sample of individual
income tax returns filed on paper. The indicator is

calculated as the percentage of refunds on those returns that are
free of any IRS- caused errors in the name and address field or in
the refund amount. g This indicator is based on a sample of paper
returns and is an average calculated starting from the

signature date on the return to the date the taxpayer should have
received the refund, allowing 2 or 3 days after issuance
(depending on whether the refund is paid by check or direct
deposit) for the refund to reach the taxpayer or the taxpayer's
bank account. h This indicator is based on a sample of
electronically filed returns and is an average calculated from

the date the return is received to the date the taxpayer should
have received the refund, allowing 2 or 3 days after issuance
(depending on whether the refund is by check or direct deposit)
for the refund to reach the taxpayer or the taxpayer's bank
account. i The level of service indicator is calculated by
dividing the number of calls answered by the total call

attempts. Calls answered includes the calls to the voice messaging
system that were subsequently returned by IRS. Total call attempts
is the sum of calls answered, calls abandoned by the caller before
receiving assistance, and calls that receive a busy signal. IRS
did not track level of service in 1997. We calculated the 1997
level of service using IRS data for that year. j IRS did not
establish goals for this indicator because it was selected as an
indicator after planning for

the 1998 filing season had been completed. Although this indicator
did not exist in 1997, we computed the 1997 accomplishment using
IRS' data and methodology. k For 1998, the level of access is the
sum of the number of calls answered plus the number of

abandoned calls divided by the total call attempts (components are
defined as in note i above). The 1997 and 1998 accomplishments
cannot be compared because IRS changed its calculation method. In
1997, the calculation was the number of calls answered divided by
the number of individual callers. In 1998, the denominator was the
number of call attempts. Therefore, if one person called five
times before reaching IRS, the denominator would have been 1 in
1997 and 5 in 1998. l This indicator measures the accuracy of tax
law information provided to taxpayers through the tollfree

telephone assistance program. It is based on test calls placed to
the telephone assistors. m The accuracy with which forms
distribution centers process taxpayers' orders is determined by

randomly checking selected taxpayer orders, monitoring telephone
calls from taxpayers, and reviewing data transcription of written
form requests from taxpayers.

Source: IRS data and GAO analysis of IRS data.

During our reviews of past filing seasons, we assessed the
methodology behind some of the indicators in table 1 and validated
some of IRS' reported accomplishments. During the 1994 filing
season, we examined IRS' methodology for testing the timeliness
and accuracy of refunds issued to taxpayers who filed paper
returns. 7 We replicated IRS' test at one service center and
concluded that it provided a valid measure of refund accuracy and
timeliness at that center. We are comfortable with IRS'
methodologies for computing the level of access and level of
service associated with its taxpayer service telephone system
because we have been working with IRS over the past few years to
develop these measures. Also, although we have not verified the
validity of the data used in those calculations, we have, in
several years, done our own tests of accessibility,

7 Tax Administration: Continuing Problems Affect Otherwise
Successful 1994 Filing Season (GAO/ GGD95- 5, Oct. 7, 1994).

B-279310 Page 8 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

which produced results that were consistent with IRS' reported
results. We discuss the results of our test for the 1998 filing
season later in this report.

Our work during the 1990 and 1991 filing seasons provided
assurance that IRS' methodology for measuring the accuracy of tax
law assistance was reliable. 8 IRS followed the same methodology
during the 1998 filing season, except that test calls were made by
a contractor rather than by IRS employees. We also have done work
in the past to determine the accuracy with which IRS' form
distribution centers process taxpayers' orders 9 and to assess a
contractor's methodology for testing the performance of IRS'
distribution centers. 10 IRS conducted this same test during the
1998 filing season using the contractor's methodology.

Among other things, IRS' walk- in sites are to distribute forms,
help taxpayers prepare tax returns, resolve account issues
detailed in notices to taxpayers, and answer tax law questions.
IRS data show that walk- in sites served about 6.2 million
taxpayers between January 1 and April 25, 1998. According to IRS'
business vision and its walk- in site mission statement, walk- in
operations are to provide accessible, high- quality service to the
public; reduce taxpayer burden; and ensure compliance with tax
laws. However, as noted in our recent report on measuring customer
service 11 and confirmed during this review, IRS' walk- in program
lacked key performance indicators for quality; timeliness; and
service accessibility. 12

IRS' National Office provided regional offices with some overall
management guidance on the walk- in program during the 1998 filing
season. However, that guidance did not include specific
requirements on how certain items, such as quality, should be
measured. To measure

8 Tax Administration: IRS' 1990 Filing Season Performance
Continued Recent Positive Trends (GAO/GGD-91-23, Dec. 27, 1990)
and Tax Administration: A Generally Successful Filing Season in
1991 (GAO/GGD-91-98, June 28, 1991).

9 Tax Administration: IRS' 1992 Filing Season Was Successful but
Not Without Problems (GAO/ GGD- 92132, Sept. 15, 1992). 10 Tax
Administration: Increased Fraud and Poor Taxpayer Access to IRS
Cloud 1993 Filing Season (GAO/GGD-94-65, Dec. 22, 1993). 11 Tax
Administration: IRS Faces Challenges in Measuring Customer Service
(GAO/GGD-98-59, Feb. 23, 1998). 12 We recognize that a complete
evaluation of the walk- in program would involve other factors,
such as the comparative costs and benefits of (1) answering tax
law questions or resolving account issues at walk- in sites versus
over the telephone, (2) return preparation at walk- in sites
versus return preparation by community volunteers, and (3) forms
distribution at walk- in sites versus communitybased locations. An
analysis of costs and benefits for these various services was
beyond the scope of this review. IRS Lacked Key

Performance Indicators With Which to Measure the Services Provided
by WalkIn Sites During the 1998 Filing Season

B-279310 Page 9 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

quality, the guidance stated that regions should do program
reviews and visit sites to ensure that consistent and accurate
service was being provided. However, the guidance did not specify
what regional officials should be examining during their reviews.

We gathered data about walk- in program reviews at three of IRS'
four regions. Two regions did such reviews but conducted the
reviews differently; the other region did a pre- filing- season
readiness review but did no reviews during the filing season. Even
though National Office guidance encouraged that the reviews be
unannounced, one of the two regions that did reviews notified the
sites of the visits in advance because, according to a cognizant
official, the guidance merely encouraged unannounced visits.
Therefore, sites in this region may have had time to prepare for
the visits and be on their best behavior. The other region that
conducted visits did those visits unannounced, with IRS staff
posing as taxpayers.

Regions were not required to report the results of any walk- in
site performance reviews to the National Office. Nevertheless, the
region that did the unannounced visits and had IRS staff pose as
taxpayers reported its results and identified several concerns.
Among those concerns were (1) incorrect advice by walk- in site
assistors concerning qualification for the EIC and (2) the failure
of assistors to probe for information to ensure that they
accurately responded to the taxpayers questions. The region also
expressed great concern about the manner in which many taxpayers
were treated and said that its observations pointed to the need
for basic training in and monitoring of customer service skills.
The other region that did reviews did not report its results.

Regarding timeliness, the National Office established taxpayer
wait- time goals of 30 minutes for return preparation and 15
minutes for all other issues. Even though the National Office
established these goals, it did not have a mechanism for
monitoring walk- in sites' performance. For example, the National
Office did not require the regions to report wait times;
therefore, two of the three regions for which we had information,
did not report any wait- time data, and the third reported such
data only to the extent that they were observed during performance
reviews. Thus, IRS cannot determine if the walk- in program met
the wait- time goals of 15 and 30 minutes during the 1998 filing
season.

IRS' Internal Audit reviewed the walk- in program for the 1997
filing season and recommended that IRS develop a customer- driven
strategy that bases

B-279310 Page 10 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

program goals and measures on customer feedback. 13 In response to
that recommendation, IRS contracted with a firm to develop and
analyze the results of a customer satisfaction survey. The survey
was designed to gather information on issues such as the
accessibility of IRS' service (e. g., the convenience of walk- in
office hours). Due to printing and distribution problems, however,
the survey was shipped to walk- in sites too late in the 1998
filing season to provide a complete assessment. According to IRS,
most sites did not receive the survey until mid- to late- March
1998.

In a report on the results of walk- in surveys done during March,
April, and May, 1998, the contractor said that Walk- in customers
as a whole indicated high overall satisfaction. Of the nine areas
that customers were asked to rate, employee courtesy and treating
you fairly scored the highest, while convenience of office hours
scored the lowest.

According to IRS, it plans to better measure walk- in performance
in 1999. It plans to measure customer satisfaction (with a survey
starting earlier in the filing season than was the case in 1998),
the number of customers served, and customer wait times. The 15-
and 30- minute wait- time goals that were in effect in 1998 are to
be retained in 1999. Also, according to IRS, quality will be
assessed through activities such as regional readiness reviews and
program reviews.

As of October 30, 1998, IRS had received 123.1 million individual
income tax returns, which was an increase of about 2 percent
compared to the same time last year. While the increase in the
overall number of returns filed was small, the use of electronic
filing increased at a much more robust pace. These increases were
consistent with filing patterns in 1997. Even with the increase in
electronic filing, about 98. 5 million tax returns (80 percent)
were filed on paper in 1998.

IRS offers three alternatives to the traditional filing of paper
returns. Two of those alternatives involve electronic filing that
is, traditional electronic filing 14 and TeleFile. 15 The third
alternative, Form 1040PC (Individual

13 Taxpayer Walk- In Program for the 1997 Filing Season, IRS
Internal Audit, Reference No. 081004, December 22, 1997. 14
Traditional electronic filing involves the transmission of returns
over communication lines through a third party (such as a tax
return preparer or electronic return transmitter) to an IRS
service center, where the return data are automatically edited and
processed. Traditional electronic filing can take several forms.
Taxpayers can (1) have their returns prepared by someone else,
such as H& R Block, and have the preparer arrange to transmit the
return to IRS; (2) take a self- prepared return to an authorized
transmitter; or (3) prepare their returns on a computer using
certain return preparation software and transmit the return on-
line to a third party who then transmits it to IRS. Use of
Electronic

Filing Continues to Increase

B-279310 Page 11 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Income Tax Return- Computer Prepared), provides another way to
file on paper. 16 Although IRS has not done the kind of
comprehensive analysis needed to fully assess the costs and
benefits associated with these alternative filing methods, it
assumes that the methods save IRS money by, among other things,
significantly reducing the number of errors that IRS has to
correct. Traditional electronic filing and TeleFile, for example,
include built- in checks that are designed to catch certain
taxpayer errors, such as computational mistakes, in advance so
that they can be corrected by the taxpayer before IRS takes
possession of the return. Also, returns filed electronically
bypass the error- prone manual procedures that IRS uses to process
paper returns.

As shown in table 2, the use of both electronic filing
alternatives increased substantially in 1997 and 1998, while the
use of Form 1040PC declined in 1998 after substantial growth in
1997.

(Number of returns in thousands)

Filing type 1/ 1/ 96 to 11/ 1/ 96 1/ 1/ 97 to

10/ 31/ 97 Percentage

change: 1996 to 1997 1/ 1/ 98 to

10/ 30/ 98 Percentage

change: 1997 to

1998

Paper Traditional

Form 1040 60,536 61,413 1. 45 61,664 0. 41 Form 1040A 18,839
17,427 -7.50 16,440 -5.66 Form 1040EZ 17,167 14,523 -15.40 12,816
-11.75 Form 1040T a 253 N/ A N/ A N/ A N/ A

Subtotal 96,795 93,363 -3.55 90,920 -2.62

Form 1040PC 7, 042 8,427 19.67 7,533 -10.61

Subtotal 103,837 101,790 -1.97 98,453 -3.28

Electronic Traditional 12,140 14,457 19.09 18,639 28.93 TeleFile
2,840 4,694 65.28 5,963 27.03

Subtotal 14,980 19,151 27.84 24,602 28.46 Total 118,817 120,941 1.
79 123,055 1. 75

Legend: N/ A = not applicable. a IRS developed Form 1040T to test
a document scanning and imaging system that was eventually

terminated in October 1996. Source: IRS' Management Information
System for Top Level Executives.

15 Under TeleFile, certain taxpayers who are eligible to file a
Form 1040EZ (Income Tax Return for Single and Joint Filers With No
Dependents) are allowed to file using a toll- free number on
Touch- Tone telephones.

16 Under the Form 1040PC method, a taxpayer or tax return preparer
uses computer software that produces a paper tax return in an
answer- sheet format. The Form 1040PC shows the tax return line
number and the data (dollar amount, name, etc.) on that line. Only
lines on which the taxpayer has made an entry are included on the
Form 1040PC.

Table 2: Number of Individual Income Tax Returns Received, by
Filing Type

B-279310 Page 12 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

One barrier to the increase of electronic filing in the past has
been its cost if taxpayers wanted to file electronically, they
generally had to pay more than they would pay to file on paper.
Two developments in the private sector reduced the significance of
cost as a barrier to taxpayers in 1998 and, according to a
cognizant IRS official, most likely contributed to the increase in
traditional electronic filing. First, the largest nationwide
return preparation company revised its fee structure to include
electronic filing in the basic return preparation fee. Thus,
unlike past years, the decision of a person who used this company
to file electronically was not influenced by cost, because it cost
the same whether he or she chose to file on paper or
electronically. Second, some commercial software packages that
were available for use in preparing tax year 1997 income tax
returns included, in the cost of the software, the ability to file
a return from a personal computer through an on- line
intermediary. That form of electronic filing accounted for about
922, 000 returns in 1998 compared to about 361,000 in 1997 an 155-
percent increase.

Another major barrier to the increase of electronic filing has
been the fact that traditional electronic filing is not completely
paperless. For example, taxpayers must send IRS their Forms W- 2
(Wage and Tax Statements) and Form 8453 (a signature document)
after their returns have been electronically transmitted. IRS must
then manually input data from these documents and match them to
the electronic return.

In 1998, as in the prior two filing seasons, IRS tested the use of
digitized signatures at a limited number of locations. In the
test, taxpayers used an electronic signature pad in lieu of
signing a Form 8453. The electronic signature was attached to the
electronic return and both were transmitted to IRS. An IRS
official responsible for the program told us that the test will
not be conducted in 1999 because previous tests have indicated
that the technology can work, if properly implemented, and that
taxpayers and preparers liked the option better than using the
Form 8453. He also said that some practitioners, especially the
larger firms, had voiced a concern about the cost of signature
pads, but that they still generally preferred the use of those
pads to the burden and cost of storing and shipping paper
signature forms.

IRS has announced that it will conduct various alternative
signature tests in 1999. 17 In one test, for example, taxpayers
will choose a personal

17 In these tests, as in the digitized signature test, IRS plans
to waive the need for participants to send their Forms W- 2 to
IRS. According to a cognizant IRS official, IRS can waive the
submission of W- 2s because there is no statutory requirement that
these forms be attached to tax returns. Traditional Electronic

B-279310 Page 13 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

identification number to use when filing electronically through
certain tax preparers.

In another move toward eliminating paper, IRS has entered into
agreements with two private sector firms that will allow certain
taxpayers to pay their income taxes by credit card in 1999.
According to IRS, it plans to test the use of credit cards in 1999
and, assuming there are no problems, expand the use of credit
cards in 2000.

Taxpayers' use of TeleFile increased by 27 percent in 1998. About
onethird of that increase can be traced to a change in procedures
that made more persons eligible to use TeleFile in 1998. The rest
of TeleFile's increased use cannot be traced to any one particular
cause.

In past filing seasons, IRS was unable to accept TeleFile returns
from taxpayers whose addresses had changed since their previous
filing. IRS could not accept these returns because there was no
method within the TeleFile system to update an address and because
IRS believed that the risk of fraud would be minimized if it
required that a refund check be mailed to the taxpayer's last
official address in IRS' records.

For the 1998 filing season, IRS contracted to use the U. S. Postal
Services' National Change of Address File to update taxpayers'
addresses before the tax package mailing labels were printed. When
taxpayers called into TeleFile, they were to confirm that the
updated address was correct. If the address was not correct, the
taxpayers were ineligible to use TeleFile.

According to IRS, because of the new address update procedure in
1998, it was able to send about 2.5 million TeleFile packages
(about 10 percent of the 24.6 million TeleFile packages mailed) to
more current addresses than it otherwise would have had in the
TeleFile system. About 500,000 of those 2.5 million taxpayers used
TeleFile in 1998 (i. e., 500,000 taxpayers who would have been
ineligible to use TeleFile without this procedural change).

In an attempt to further increase the use of TeleFile, IRS, in
1999, plans to conduct a joint federal and state TeleFile test in
Indiana and Kentucky. After completing the federal portion of
TeleFile, these taxpayers are to be given the option to file their
state tax returns. If the taxpayers choose to do so, they then are
to hear instructions for filing the state return. IRS decided to
conduct this federal and state test in response to the results of
a survey of TeleFile nonusers in 1997. In responding to the
survey, 44 percent of the TeleFile

B-279310 Page 14 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

nonusers said that they might be encouraged to use TeleFile if
they could also use it to file their state returns.

In response to the same survey, 72 percent of the TeleFile
nonusers indicated that they would be encouraged to use the system
if they were able to determine their federal tax before calling
TeleFile. This would require including the tax table in the
TeleFile package a step that IRS opposes because it would increase
the size of the package by several pages (at an estimated cost of
about $1.4 million) and because the table is not needed to file a
return using TeleFile (TeleFile computes the tax for the taxpayer
on the basis of data that the taxpayer enters into the system). 18
However, given the large percentage of nonusers who indicated that
having the table would influence them to use TeleFile, some kind
of limited test, similar to what IRS is doing with respect to the
filing of state returns, might indicate whether the potential
increase in TeleFile use justifies the additional cost of
including the tax table in the package.

As of October 30, 1998, IRS had received about 7.5 million Forms
1040PC, down 11 percent from the same time in 1997. IRS data show
that, in both years, more than 75 percent of these forms were
prepared by tax return preparers. According to the largest user of
Form 1040PC in the preparer community, the drop in the number of
Forms 1040PC may be attributable to the increase in electronic
filing. This tax return preparation company was the one referred
to previously that included the cost of electronic filing in its
return preparation fees in 1998. According to data provided by the
firm, more clients chose to have their returns filed
electronically in 1998 compared to 1997 and fewer Forms 1040PC
were filed.

In our reports on the 1996 and 1997 filing seasons, we commented
on (1) IRS' use of lockboxes, which are postal rental boxes
serviced by commercial banks, to process tax payments submitted
with Forms 1040 and (2) IRS' requirement, as part of that
arrangement, that taxpayers mail to the banks not only their tax
payments but their tax returns. IRS and FMS assume that the use of
lockboxes is beneficial to the government because, in general,
banks can process the payments and deposit the money to a Treasury
account quicker than service centers. This means that Treasury
would not have to borrow as much to pay government obligations,
thereby avoiding interest charges.

18 IRS' survey did not solicit information that would explain why
taxpayers wanted to be able to determine their federal tax before
calling TeleFile. However, one possible reason might be their
desire to independently compute their tax liability so that they
would have some idea of what to expect before calling TeleFile.
Form 1040PC

IRS Still Has Inadequate Evidence to Determine the Most
Appropriate Lockbox Procedure

B-279310 Page 15 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

After reviewing information compiled by IRS and FMS in 1998, we
continue to believe that IRS has inadequate evidence to determine
whether the additional burden that may be caused by having
taxpayers separate their returns from their payments outweighs the
several million dollars in additional cost associated with having
the banks handle those returns. Nevertheless, IRS plans to
continue the current lockbox process in 1999.

In our report on the 1996 filing season, we questioned IRS'
decision to have taxpayers send both their tax returns and their
tax payments to lockboxes. 19 The banks do not need tax returns to
process remittances, and the government was incurring several
million dollars in additional cost by requiring the banks to
receive returns, sort them, and ship them to IRS for processing
even after netting out what it would cost if IRS sorted the
returns instead of the banks. IRS believed that the alternative
asking taxpayers to separate their tax payments from their returns
and mail each to a different address would impose a burden on
taxpayers. In our opinion, the evidence IRS provided in 1996 on
taxpayer burden was not convincing. We recommended that the
Commissioner of Internal Revenue either discontinue having returns
sorted by the banks, thus reducing the extra cost to the
government, or reconsider the decision to have taxpayers send
their tax returns to the banks along with their tax payments.

In response to our recommendation, IRS said that it had formed a
task force to identify a long- term solution for directing Form
1040 tax payments to lockboxes. In May 1997, the task force
recommended that IRS have taxpayers separate their returns from
their payments. Despite the task force's recommendation, IRS
decided that lockboxes would continue to receive and sort tax
returns because of IRS' continuing concern about taxpayer burden.
To support its position, IRS cited the results of several focus
groups that became available after the task force had completed
its work.

In our report on the 1997 filing season, we concluded, after
reviewing the results of those focus groups, that IRS still did
not have conclusive evidence to support its position. 20 We also
questioned the validity of IRS' and FMS' assumption that lockboxes
can process remittances and deposit the money to a Treasury
account 3 days faster than service centers. That assumption was
key to the assertion that it is cost beneficial to have lockboxes
process Form 1040 tax payments. We recommended, among

19 GAO/GGD-97-25. 20 GAO/GGD-98-33.

B-279310 Page 16 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

other things, that IRS (1) conduct, during the 1998 filing season,
the analyses necessary to determine whether there are net savings
to the government attributable to the use of lockboxes to process
Form 1040 tax payments and (2) collect definitive data on whether
taxpayers believe, given the processing cost savings to the
government, that it would cause them an unreasonable burden to
mail tax returns and tax payments to different locations. We also
recommended that IRS take certain steps depending on the results
of the analyses.

In an August 1998 report, Treasury's Office of the Inspector
General concurred with our findings and also recommended that IRS
acquire relevant and reliable comparative cost data on all aspects
of the lockbox program to identify the most cost- effective option
to use when the current lockbox arrangement expires in 1999. 21

In 1998, in response to our recommendations and in line with the
Inspector General's recommendation, (1) FMS contracted for a study
to determine how much time the government saves, if any, by having
lockboxes instead of service centers process remittances and (2)
IRS included lockboxrelated questions in 10 focus groups involving
about 100 taxpayers.

To determine how much time the government saves, if any, by having
lockboxes instead of service centers process remittances, FMS, in
March 1998, contracted for a study that was designed to measure
and compare processing time frames. The contractor was to analyze
the following three factors: (1) the average mail time to a
processing site, (2) the average elapsed time from the receipt of
mail at the lockbox or service center to the time receipts were
deposited in the bank, and (3) the average time required to
convert deposited receipts to funds available in the Treasury
account.

Because we did not receive a copy of the contractor's July 1998
report until October 1998, we were only able to partially assess
its implications for this report. The study showed that, on
average, lockboxes process remittances about 2 days faster than
the service centers 1 day less than both FMS and IRS had been
assuming in calculating the interest cost avoidance from using
lockboxes. The contractor also made several observations about
existing processes that, if changed, might reduce service center
processing times. For example, the contractor noted that
processing times might be reduced if (1) service centers had more
reliable

21 Review of the Effectiveness of Using Commercial Bank Lockboxes
for Federal Income Tax Payments, Department of the Treasury,
Office of Inspector General, Reference No. OIG- 98- 097, August
20, 1998. FMS Study

B-279310 Page 17 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

mail- opening equipment, (2) the strategies at service centers for
working on weekends and making deposits on the weekend were
consistent, and (3) IRS found some way to better identify
envelopes that contained remittances like the different- colored
mailing labels that several states use to distinguish tax returns
with payments from returns that involve a refund.

Using the average 2- day timeline variance and data provided by
FMS on the then current value of funds rate (5 percent) and the
amount of Form 1040 remittances collected by the banks in fiscal
year 1998 (about $48.3 billion), we determined that the lockbox
processing of Form 1040 remittances resulted in an interest cost
avoidance of $19.3 million. 22 In addition, a cognizant FMS
official told us that the contractor's study indicated that the
timeline variance was actually 3- days during parts of various
peak processing periods because of expedited weekend processing by
lockboxes. Assuming that the 3- day variance for certain periods
is correct and using data provided by FMS, we determined that the
amount of interest cost avoidance should be increased by about
$1.6 million, for a total of about $20.9 million. 23 Other data
obtained from FMS on the number of items handled by the banks
(about 11 million) and the banks' charges per item show that the
government paid banks about $19.8 million for processing the Form
1040 remittances in fiscal year 1998, of which about $9.7 million
was to reimburse the banks for sorting tax returns and shipping
them to IRS.

Taken together, and assuming the appropriateness of the 3- day
variance for certain weekend processing, FMS' data indicate that
the use of lockboxes to process Form 1040 remittances saved the
government about $1.1 million in fiscal year 1998 ($ 20.9 million
in interest cost avoidance minus $19.8 million in bank charges).
However, the data also indicate that the use of lockboxes could
save the government substantially more if the banks were not
required to handle tax returns. For example, according to FMS, the
$9.7 million paid to the banks for handling tax returns included
74 cents a return for sorting and 13 cents a return for shipping.
Two years ago, IRS told us that having service centers sort the
returns would cost IRS

22 FMS used the average variance of 2 days to compute interest
cost avoidance. However, looking at each of the 10 service centers
and its corresponding lockbox bank separately, the timeline
variances reported by the contractor ranged from about 3/ 4ths of
a day to 3- 1/ 2 days. We do not know how much, if at all, the
result would have changed if FMS had calculated a separate
interest cost avoidance for each of the 10 locations and then
summed those 10 calculations.

23 Although we could confirm that the contractor identified
differences between banks and at least some service centers with
respect to weekend processing, we did not have time to determine
whether the contractor's data supported FMS' decision to use a 3-
day variance for certain processing periods.

B-279310 Page 18 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

an average of 37 cents a return one- half as much as the 74 cents
charged by the banks. Thus, assuming that IRS' sorting costs
remained at 37 cents a return, 24 having the banks sort and ship
11 million returns in fiscal year 1998 would have cost the
government an additional $5.5 million (11 million returns times 50
cents of additional cost per return that is, 37 cents for sorting
and 13 cents for shipping).

In an attempt to obtain information on taxpayer burden associated
with having taxpayers mail their tax returns and tax payments to
separate locations, IRS added questions to an existing focus group
study. That study, done by a contractor, was originally designed
to solicit ideas from taxpayers on ways to improve the Form 1040
tax package. IRS added questions to solicit taxpayers' views on
mailing their returns and remittances in separate envelopes. The
contractor held 10 focus groups involving about 100 taxpayers who
prepared their own federal income tax returns.

Although they provided some evidence of taxpayer concern about
burden, these focus groups did not, as we had recommended in 1997,
provide definitive data on whether taxpayers believe, given the
processing cost savings to the government, that it would cause
them unreasonable burden to mail tax returns and tax payments to
different locations. Most important, as noted in our report on the
1997 filing season, although focus groups are useful in providing
insight on a particular issue, they are not statistically
representative of the population and should not, in and of
themselves, provide the basis for far- reaching conclusions. As
noted in the contractor's report, its study was qualitative in
nature and the hypotheses discussed in this report should be
viewed as tentative.

Also, data generated by the focus groups did not provide
definitive evidence that taxpayers considered the increased burden
associated with using two envelopes unreasonable. The concerns
mentioned by participants involved (1) the need to use two
envelopes instead of one if they owed money, 25 (2) the need to
use an additional postage stamp if they had to use two envelopes,
and (3) the possibility that payments and returns would not get
linked if they were sent to different places.

24 We do not know what IRS' per- return sorting costs were in
1998. Although some factors, such as increased salaries and
benefits, could have driven those costs up, other factors, such as
improved productivity or decreased overhead, could have had the
opposite effect.

25 For persons using a tax package, the two envelopes would be
provided by IRS and included in the package. Persons not using a
package would have to supply their own envelopes. IRS Focus Groups

B-279310 Page 19 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

It is true that taxpayers would have to use two envelopes instead
of one and would have to use an extra stamp, but there is no
indication in the contractor's May 1998 summary report that focus
group participants were told about the potential savings to the
government if taxpayers assumed the additional burden. In that
regard, the contractor's report noted that some taxpayers, who
imagined the system would save the IRS time or money, said they
would be willing to go along with the new idea if it was properly
explained to them. In our opinion, participants should have been
told about potential savings rather than having to imagine them.

As for the participants' concern about separating their returns
from their payments and the possibility that the two would not get
reconnected, there was no indication in the report that
participants were told that the current procedure also results in
the payment and return being separated. When taxpayers mail their
returns and payments to a lockbox, the bank separates the payments
from the returns and ships the returns to IRS. The bank then
processes the payments while IRS processes the returns.

Given all of the previously mentioned information, we believe that
IRS still has not obtained definitive evidence showing that the
additional taxpayer burden and costs associated with the use of
two envelopes would outweigh the additional costs the government
incurs to have banks sort and ship tax returns.

Responsible IRS officials told us that IRS will continue the
current lockbox process in 1999. According to the officials, IRS
decided to continue the current program because FMS' study was not
finalized in time to make changes for 1999 and because IRS
continues to believe that asking taxpayers to separate their
returns and payments would impose an unreasonable burden. However,
in commenting on a draft of this report, IRS said that additional
analysis is warranted regarding the efficacy of, and alternatives
to, the current arrangement. In that regard, IRS said that it
would do a detailed analysis of the costs and feasibility of
returning this portion of the remittance processing workload to
IRS at some point in the future. IRS' Plans for 1999

B-279310 Page 20 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

The Taxpayer Relief Act of 1997 (TRA97) changed the way capital
gains are treated for income tax purposes. Before the act, capital
gains were taxed at the same rate as other income, up to a maximum
of 28 percent. TRA97 created capital gains tax rates ranging from
10 percent to 28 percent, depending on the type of asset, the
holding period of the asset, the date of sale, and the
individual's income. 26 As a result, taxpayers must now report to
IRS more capital gains information than previously required. To
implement the new legislation, IRS revised Schedule D (Capital
Gains and Losses). That revision increased the number of lines on
the Schedule D. The schedule used in 1997 had 19 lines and a 13-
line worksheet; the revised schedule had 54 lines.

TRA97 was not passed until August 5, 1997, and some issues related
to capital gains were not resolved until early October 1997.
Because of the time needed to make the necessary computer software
changes, IRS was not prepared to process returns with Schedule Ds
until February 13, 1998. Until then, IRS could not accept
electronic returns with Schedule Ds, and service centers had to
suspend the processing of paper returns with Schedule Ds.

A representative of one of the tax return preparation firms we
contacted told us that IRS' hold on accepting electronic returns
with Schedule Ds required his firm to develop a software program
that would hold electronic returns with Schedule Ds in suspense
within its computer system. This firm had over 100,000 electronic
returns with Schedule Ds in suspense that had to be transmitted
once IRS announced it could process Schedule Ds.

Officials at the 5 service centers we visited said that their
inventories of suspended paper returns with Schedule Ds were not
very great (fewer than 5,000 at each service center) and were
cleared quickly after the software changes were made. According to
the officials, the impact was not significant because,
historically, most returns with a Schedule D come in later in the
filing season.

According to an IRS analysis of a statistically valid sample of
tax year 1997 returns filed through August 28, 1998, the
percentage of individual income tax returns filed with a Schedule
D increased from 13 percent in 1997 to 19 percent in 1998. This
increase was partly due to the requirement that all capital gain
distributions from mutual funds be reported on Schedule D. In past
years, taxpayers with capital gain distributions from mutual funds
under $400, and no other capital gains or losses, were not
required to

26 Normal marginal rates for taxable income range from 15 percent
to 39.6 percent. Implementation of

Change to Capital Gain Provisions Led to Some Processing Delays
and Increased Taxpayer Burden

B-279310 Page 21 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

complete a Schedule D. The IRS analysis noted that the number of
individual income tax returns filed in 1998 that showed capital
gain distributions increased by over 3 million compared to the
number in 1997.

IRS estimates that about 1 million taxpayers with capital gains
failed to submit a Schedule D during the 1998 filing season.
Service centers were instructed to correspond with all taxpayers
who claimed capital gains of over $1 but who failed to submit a
Schedule D and to suspend the processing of those returns until
the Schedule Ds were provided. IRS officials said that they
corresponded with all of these taxpayers about the missing
Schedule Ds both to educate them about the requirement to attach a
Schedule D and because use of the Schedule D and the new capital
gain rates could reduce a person's tax liability.

In general, the tax rates on capital gains are lower than the
rates on other income. To take advantage of these lower rates,
taxpayers must compute their tax using a Schedule D. Because the
capital gains tax rates are often lower than the rates on other
income, filing a Schedule D usually reduces a taxpayer's
liability. However, for taxpayers with a small amount of capital
gains, the possible saving is minimal. By requiring everyone to
submit a Schedule D, even if they only had $1 in capital gains,
IRS created additional work for itself; delayed the processing of
some returns and the associated refunds, if any, until receipt of
the schedule; and increased taxpayer burden by causing all of
these taxpayers to complete a complicated schedule.

We reviewed a proof copy of Schedule D for use during the 1999
filing season, and the form will be virtually the same as the one
used in 1998. However, IRS has revised the way it will process
returns with missing Schedule Ds for the 1999 filing season. Under
its new procedure, IRS will process tax year 1998 returns claiming
capital gains even if there is no Schedule D attached. In doing
so, IRS will assume a short- term capital gain rate and mail a
notice to the taxpayer after the return is processed. The notice
is to inform taxpayers that their returns were processed using the
short- term rate to avoid any delay, but that they may be able to
lower their tax if long- term capital gain rates apply. The
mailing is to include a Schedule D, instructions for completing
the schedule, and a Form 1040X (amended return). Taxpayers then
will have the opportunity to file an amended return if they
determine that their tax savings are worth the additional time
needed to prepare and file the Form 1040X and Schedule D. IRS'
revised procedures should benefit many taxpayers by providing some
refunds earlier and by saving them the time associated with
completing a

B-279310 Page 22 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Schedule D if they determine that the tax savings are not
significant enough to warrant the effort.

The EIC is a refundable tax credit available to low- income,
working taxpayers. Congress created the EIC to offset the impact
of Social Security taxes and to encourage low- income workers to
seek employment rather than welfare. Because of concerns about
noncompliance and fraudulent claims, IRS has given increased
attention to EIC claims during the past few filing seasons. In
1998, IRS (1) continued validating Social Security numbers (SSN)
used in conjunction with EIC claims, (2) conducted indepth reviews
of certain EIC claims, (3) established procedures for denying
future EIC claims by taxpayers who were found to have negligently
or fraudulently claimed the EIC in 1998, (4) established
requirements for paid preparers to exercise due diligence in
determining a taxpayer's eligibility for the EIC, and (5) expanded
telephone and walk- in assistance for EIC claimants. Two main
impetuses were behind this increased activity in 1998 TRA97, which
included several provisions relating to the EIC, and a new EIC
compliance initiative that Congress began funding in fiscal year
1998.

We discussed IRS' actions in a July 1998 report to the Chairman of
the House Committee on Ways and Means and Senator Larry E. Craig.
27 As noted in that report, we believe that IRS could have done
more to alert taxpayers to the consequences of falsely claiming
the EIC and could have provided special EIC assistance earlier in
the filing season. In addition, our review of EIC processing for
the 1998 filing season disclosed that IRS does not have controls
in place to ensure that field locations receive notification of
procedural changes and implement the changes at the same time. The
lack of controls leaves open the possibility that criteria were
inconsistently applied during the processing of EIC claims in
1998.

Notwithstanding the issues discussed in our previous report and
the lack of controls identified during this review, there is
evidence to suggest that IRS' efforts have had a positive effect.
IRS data on the number of errors identified on returns with EIC
claims indicated that although taxpayers and return preparers were
still making many errors, the error rate dropped significantly in
1998. For the period January 1 through August 29, 1998, IRS
identified 1,992,379 EIC- related errors on 19,393,098 returns an
error rate of 10.27 percent. That error rate compares favorably
with 2,655,524 errors on 19,032,043 returns during the same period
in 1997 an error rate of 13.95 percent. During the same period,
there was also a significant drop in

27 GAO/GGD-98-150. IRS Expanded Its

Efforts to Ensure That Taxpayers Filed Correct EIC Claims

B-279310 Page 23 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

the number of errors involving SSNs and other taxpayer
identification numbers on EIC returns 600,676 in 1998 compared
with 955, 542 in 1997.

Since 1997, IRS has had the authority to disallow, through its
math error program, any deductions and credits associated with an
invalid SSN. 28 Under its math error program procedures, if IRS
identifies an invalid SSN while processing a return, it can
immediately adjust the taxpayer's tax liability; refund; and/ or
EIC claim. The taxpayer is to receive a notice explaining the
change to the return. Taxpayers who receive such a notice can call
or write IRS to provide the correct SSN or otherwise resolve the
issue. If taxpayers do not respond to IRS' notice, there is to be
no further correspondence unless the taxpayers fail to pay any
additional tax that was assessed as a result of an IRS change.

As of the end of August 1997, IRS had issued 955,542 math error
notices to taxpayers who had claimed an EIC but had not supplied a
correct SSN or other identification number for either themselves,
their spouses, or their dependents. As of the same point in time
in 1998, the number of such notices dropped to 600,676
representing about a 37- percent decrease from 1997. This drop in
the number of notices may indicate that fewer taxpayers are
attempting to claim an EIC to which they are not entitled. It may
also reflect the impact of IRS efforts to alert taxpayers who had
used invalid SSNs in prior years of the need to correct those
problems before filing their returns in 1998. For example:

 In December 1997, IRS sent notices to about 600,000 taxpayers who
had filed returns with an invalid SSN for the primary taxpayer. 29
The notice told the taxpayers what to do to correct the situation
before filing their returns in 1998.

 IRS also identified about 225,000 EIC- qualifying child SSNs that
had been used by more than 1 taxpayer on returns filed in 1997. In
December 1997, IRS sent notices to these taxpayers (about 383,000)
informing them of the problem and reminding them to file a correct
return in 1998.

According to IRS, its analyses of the impact of the notices mailed
in December 1997 will not be completed until sometime in 1999.

IRS identified a programming- related problem early in the 1998
filing season that raised concerns that (1) some invalid SSNs were
not being

28 IRS considers an SSN invalid if it is missing from the return
or if the SSN and associated name on the return do not match data
in the Social Security Administration's records. 29 The primary
taxpayer is the taxpayer filing the return or, on a joint return,
the taxpayer listed first. IRS Continued Validating

SSNs

B-279310 Page 24 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

resolved and (2) EIC eligibility was not being scrutinized in some
cases. Efforts to resolve the programming problem revealed that
there were no controls in place at the National Office to ensure
that service centers receive notification of procedural changes,
nor were there any standardized procedures for the service centers
to follow to ensure that the functional areas that should receive
notification of procedural changes do in fact receive those
notifications.

On January 22, 1998, IRS' National Office issued alert
instructions to the field offices to deal with the programming
problem. A cognizant program analyst in IRS' National Office told
us that the National Office does not require service centers to
respond that they have received or implemented alerts.
Representatives we contacted at three service centers told us that
the alert was received timely at each of their centers, but each
center reported a different date that the alert was received in
the branches that were to follow the instructions. The branches
received the alert on February 4, March 17, and April 28.
Therefore, because not all service center branches had received
the instructions at the same time, EIC criteria may not have been
applied consistently. IRS did not have data on the extent or
impact of this potential problem.

In April 1997, IRS published the results of its tax year 1994 EIC
compliance study. The study showed that of the $17.2 billion in
EIC claimed during the study period, taxpayers overclaimed about
$4.4 billion, or about 26 percent. In response to some of the
noncompliance problems identified in that study, IRS targeted
certain types of EIC claims for in- depth review in 1998.

One form of EIC noncompliance that IRS targeted was the use of a
qualifying child's SSN on more than one tax return for the same
tax year. As discussed in our recent report on IRS' efforts to
reduce EIC noncompliance, 30 IRS allocated staff to audit as many
as 140,000 taxpayers who had used about 92,000 duplicate
qualifying child SSNs in both tax years 1995 and 1996. 31
According to IRS officials, as of May 16, 1998, about 103,000 of
the 140,000 taxpayers had filed tax year 1997 returns, and IRS had
frozen their refunds. As of that same date, however, IRS had
released 49,000 of the frozen refunds for taxpayers who had
responded to IRS'

30 GAO/GGD-98-150. 31 The 140,000 taxpayers selected for audit as
part of this project were not included in the group of 383, 000
taxpayers previously discussed that was sent notices in December
1997 related to duplicate SSN use. IRS Targeted Certain EIC

Claims for In- Depth Review; Results Are Uncertain

B-279310 Page 25 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

correspondence on these cases but whose conflicting claims for the
child( ren) in question were not resolved. 32

IRS officials told us that the 49,000 refunds were released
because IRS (1) did not have enough time to adequately prepare for
the start of this project (e. g., get staff assigned, procedures
developed, and training done) and (2) had underestimated the
volume of contacts it would receive from taxpayers. Although IRS
is continuing to investigate these cases, its effectiveness in
protecting the revenue has been compromised because it is more
difficult (and more costly) to recoup an erroneous refund once it
has been released. As of August 21, 1998, according to IRS, it had
finished auditing about 81,500 returns under this project and had
recommended changes totaling about $63.7 million. As of that same
date, audits of another 195,900 returns were ongoing, and 77,100
returns were in inventory but not yet assigned for audit. 33

The largest source of EIC noncompliance found in IRS' tax year
1994 study related to the eligibility of qualifying children. IRS'
study further showed that a large proportion of qualifying child
errors occurred in tandem with erroneous claims of head- of-
household filing status. In that regard, another area IRS targeted
for in- depth review involved EIC claimants who filed as head of
household and whose returns contained indications of potential
qualifying child problems. IRS data show that IRS had finished
auditing about 98,500 returns under this project as of August 21,
1998, and that 75 percent of those audits had resulted in
recommended changes totaling about $169. 5 million (the other 25
percent resulted in no change to the taxpayers' returns). Also as
of August 21, about 232,800 returns were in various stages of
audit and about 33,700 returns were in inventory but not yet
assigned for audit.

Misreported income also accounted for a significant portion of the
EIC noncompliance identified in IRS' tax year 1994 study. Because
income level affects the amount of an EIC, IRS was particularly
concerned with taxpayers, such as those who are self- employed,
whose income cannot be verified through traditional compliance
activities, such as matching the amount of income reported on
information documents with income reported on tax returns. IRS
selected a sample of tax year 1997 returns

32 Also as of May 16, 1998, IRS had released refunds for 21, 000
taxpayers who did not claim the disputed child( ren) for 1997 and
another 4, 500 taxpayers who, on the basis of audits of their tax
year 1996 returns, were entitled to claim the disputed child(
ren).

33 The number of returns in this project exceeds the number of
taxpayers previously cited because the project involves multiple
tax years 1995, 1996, and 1997.

B-279310 Page 26 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

with self- employment income reported on a Schedule C (Profit or
Loss from Business) and held the refunds pending completion of the
audits. As of October 1998, according to IRS, some of these audits
had not been completed, and IRS had not compiled data on the
results of the audits that had been completed.

TRA97 provides that taxpayers who fraudulently claimed the EIC
beginning with tax year 1997 (i. e., returns filed in 1998) would
be denied the credit for the next 10 years, and that those who
negligently claimed the credit (through reckless or intentional
disregard of the regulations) would be denied the credit for the
next 2 years. When taxpayers are found to have filed fraudulently
or negligently, they must recertify eligibility after the 10- year
or 2- year sanction period to claim the EIC again.

During the 1998 filing season, IRS used its Questionable Refund
Program and other special projects to identify taxpayers who
negligently or fraudulently claimed the EIC. When fraud or
negligence is suspected, the refund is to be frozen and an in-
depth audit of the return is to be conducted a process that can
take several months to complete. After the audit is completed and
it is determined that the taxpayer fraudulently or negligently
claimed the EIC, an indicator is to be added to the taxpayer's
file to alert IRS that the taxpayer is not entitled to the EIC in
subsequent years.

IRS attempted to warn taxpayers about the implication of these new
provisions before they filed their returns in 1998 by including a
brief statement in the Form 1040 instruction about the possibility
that taxpayers may not be allowed to claim the EIC in the future.
Although we have no way of knowing how successful that warning was
in encouraging better compliance, we believe that the chances for
success might have been enhanced if IRS had done a better job of
publicizing the warning. In our July 1998 report, 34 we
recommended that information regarding the 2- year and 10- year
sanctions and the recertification process be prominently published
in the Form 1040 EIC instructions and on the Schedule EIC.

In commenting on our recommendation, IRS said that it would
include guidance in the tax year 1998 Schedule EIC instructions to
advise taxpayers of the 2- year and 10- year sanctions and the
need to provide IRS with additional information when they file for
the credit in a subsequent year. However, IRS said that it did not
intend to revise the Schedule EIC because these issues do not
involve the majority of filers. IRS believes that

34 GAO/GGD-98-150. New Procedures for

Negligent or Fraudulent Disregard of EIC Rules Could Have Been
Better Publicized

B-279310 Page 27 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

providing the information on the schedule may confuse filers who
have not had their credit disallowed. Also, according to IRS,
taxpayers must go to the instructions to determine if they qualify
for the credit before completing the schedule, and IRS' intent is
to place the information so that persons using the instructions
will easily see it.

In evaluating IRS' comments in our July 1998 report, we said that
although inclusion of information in the Schedule EIC instructions
would be an improvement, we continued to believe that something
should also be added to the schedule. Because one of the purposes
of this information is to alert potential EIC claimants to
possible repercussions if they make erroneous claims, the
information affects more filers than those who have had their EIC
claim disallowed. Also, although it is true that taxpayers who
choose to compute their own EIC have to use the worksheet in the
instructions, taxpayers who choose to have their returns prepared
by someone else do not have to use the worksheet and thus might
see only the Schedule EIC. A brief, but prominent, cautionary
statement added to the Schedule EIC would alert those taxpayers to
important information in the instructions that they should read
before filing their returns.

TRA97 also imposes an $100 penalty on paid tax return preparers
who fail to exercise due diligence in determining a taxpayer's
eligibility for the EIC. In December 1997, IRS issued specific due
diligence requirements and publicized them in mailings to
practitioners. Under these requirements, a paid preparer must (1)
complete an EIC eligibility checklist or a substitute form that
contains the same information; (2) complete the EIC worksheet or
keep a paper or electronic record of the EIC computation that
includes the computation method and information used; (3) not know
or have reason to know that any information used to determine EIC
eligibility is incorrect; and (4) retain, for 3 years, a copy of
the completed checklist, the worksheet, and a record of how and
when the information was obtained and who provided the
information.

We did not assess the impact of these due diligence requirements
on preparers or taxpayers. However, a representative of one of the
tax return preparation firms we contacted told us that it was not
difficult to comply with the requirements and that going through
the EIC eligibility checklist makes it easier to interview clients
to make sure they are entitled to the EIC.

IRS did not institute at a national level specific procedures to
monitor compliance with the due diligence requirements during the
1998 filing season, but we know of at least one field office that
did some monitoring. IRS Established

Requirements for Paid Preparers to Exercise Due Diligence in
Determining EIC Eligibility

B-279310 Page 28 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

IRS told us that its national- level plans for the 1999 filing
season include due diligence monitoring visits to EIC return
preparers, but IRS has not yet decided on the procedures for these
visits, the number of visits, or the extent to which IRS will
target those preparers most likely to be noncompliant.

In an effort to help taxpayers prepare more compliant EIC claims
in 1998, IRS, among other things, expanded telephone access for
EIC- related issues to 7 days a week, 24 hours a day. Also,
starting on March 7, 1998, IRS provided assistance on 6
consecutive Saturdays (when offices are normally closed) at more
than 150 walk- in sites. One of those Saturdays (Mar. 28) was
promoted as EIC Awareness Day.

According to IRS data, 95,000 taxpayers called the EIC assistance
lines during the times when IRS' other assistance lines were not
available. IRS data also show that walk- in staff helped 2,949 EIC
claimants prepare their returns on Saturdays and provided 1,032
others with different types of EICrelated assistance. According to
IRS, this is in addition to 185,305 EIC claimants who were
assisted on weekdays during the filing season.

Although many taxpayers who were eligible for the EIC were helped
by these efforts, we believe more taxpayers could have benefited
had IRS implemented its assistance differently. For example:

 IRS did not advertise the 24- hour availability of telephone
assistance for EIC- related issues. IRS informed taxpayers of this
service only if they received a notice from IRS about a problem
with the EIC claims on their tax returns. IRS officials told us
that they did not advertise this service because they thought that
it would lead to many non- EIC calls during the hours when other
assistance lines were closed. 35

 By March 7, 1998, when IRS first offered Saturday assistance at
its walk- in sites, millions of EIC claims had already been filed.
IRS said that it did not offer Saturday service earlier in the
year because prior to receiving the appropriation [for the new EIC
compliance initiative], we had anticipated having Saturday service
for only the last six weeks of the filing season when, according
to IRS officials, demand among all filers is generally higher.
Also according to IRS, the date for EIC Awareness Day was selected
so that IRS would have adequate time to publicize and provide for
quality service to the public.

35 This should not be a problem in 1999, when IRS plans to offer
24- hour assistance on all of its telephone assistance lines.
Expanded Telephone and

Walk- In Service for EIC Claimants Could Have Been More Beneficial

B-279310 Page 29 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

In our July 1998 report, 36 we recommended that IRS ensure that
customer service efforts aimed at EIC claimants are available
earlier in the filing season. IRS agreed with our recommendation
and said that it would (1) advertise the availability of 7- days-
a- week, 24- hours- a- day telephone assistance (for tax law
questions, questions on notices, and other issues) in all tax
packages for the 1999 filing season and (2) offer Saturday walk-
in service from January 16 to April 10, 1999, with the first 6
Saturdays designated as EIC Awareness Days.

During each filing season, millions of taxpayers call IRS to ask
questions about the tax law, their refunds, or their accounts and
to order forms. The extent to which taxpayers are able to get
through to IRS when they call is an important indicator of filing
season performance. IRS performance indicators show that
taxpayers' ability to reach IRS by telephone increased
significantly during the 1998 filing season. 37

In 1998, IRS used two measures to gauge taxpayers' success in
getting through to IRS over the telephone level of access and
level of service. 38 The only difference between the two measures
is that level of access is computed by including abandoned calls
in the number of calls that gained access (i. e., the callers had
gained access to IRS' system but subsequently decided, for unknown
reasons, to hang up before an assistor came on the line) while
level of service is computed after deleting the abandoned calls
(i. e., although those callers gained access to the system, they
were not served).

According to IRS data, as shown in table 3, level of access
increased by 26.1 percentage points (from 64.5 percent in 1997 to
90.6 percent in 1998) and level of service increased by 21.1
percentage points (from 52.6 percent in 1997 to 73.7 percent in
1998). 39

36 GAO/GGD-98-150. 37 For purposes of toll- free accessibility,
the 1998 filing season was from January 1 to April 18, 1998, and
the 1997 filing season was from January 1 to April 19, 1997. 38
Level of access is defined as the sum of the number of calls
answered and the number of calls that are abandoned by the caller
before getting assistance divided by total call attempts (which
consist of calls answered, calls that are abandoned, and calls
that receive a busy signal). Level of service is defined as the
number of calls answered divided by total call attempts.

39 In reporting 1998 telephone data, IRS combined data on six of
its toll- free telephone lines tax law assistance, EIC/ refund
inquiry, account inquiry, forms ordering, Automated Collection
System, and fraud hotline. In reporting 1997 telephone data, IRS
did not include one of those six lines the line for the Automated
Collection System. Therefore, to compare IRS' performance during
the 2 years, we omitted information for the Automated Collection
System from all data calculations and presentations. Access to
IRS' Phone

Systems Improved

B-279310 Page 30 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Filing season a Call

attempts (millions)

Calls abandoned

(millions) Calls

answered (millions)

Level of access (percent)

Level of service (percent)

1997 70.2 8. 3 36.9 64.5 b 52.6 1998 48.5 8. 2 35.7 90.6 73.7
Change (21.7) c (. 1) (1.2) 26.1 21.1 a Data are for January 1
through April 19, 1997, and January 1 through April 18, 1998.

b We calculated the level of access for 1997 by using IRS' current
methodology, which is based on the number of call attempts, not
the methodology IRS used during 1997, which was based on the
number of callers. Therefore, this percentage does not match the
percentage shown in table 1 (see table 1, note k). c The
significant reduction in call attempts was apparently due, in
large part, to the 82- percent

decrease in the number of busy signals, which is discussed later
in this report. Source: GAO analysis of IRS data.

During the 1997 and 1998 filing seasons, we conducted tests, using
judgmental samples, of the accessibility of one of IRS' toll- free
telephone lines the line taxpayers are to call if they need tax
law assistance. The results of those tests also indicated that
taxpayers were better able to contact IRS over the telephone in
1998 compared to 1997. Although our test results cannot be
projected to all calls made to IRS, we did take steps to ensure
that a random calling pattern was used to negate that source of
potential bias in our tests.

Results of our test during the 1998 filing season showed a level
of access of about 81 percent and a level of service of about 75
percent. Results of a similar test we conducted during the 1997
filing season showed a 51percent level of access and a 39- percent
level of service. Our 1998 test methodology and detailed results
are described in appendix I.

Several factors may have contributed to the improved rates of
telephone access and service reported by IRS and indicated by our
tests. Specifically, IRS (1) extended the hours during which
taxpayers could call IRS and speak to an assistor, (2) expanded
its use of voice messaging, (3) increased the number of staff
available to answer the telephone during peak calling periods so
that more calls were answered on the first attempt by the
taxpayer, (4) offered taxpayers who were on hold more than 10
minutes the option of leaving a recorded message and receiving a
return call, and (5) improved its system of routing calls to
available call sites around the country.

Table 3: Accessibility of IRS' Telephone Assistance

B-279310 Page 31 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

IRS increased the hours of service on the toll- free telephone
system to 16 hours a day, 6 days a week. 40 As a result, taxpayers
seeking assistance via the toll- free telephone system were able
to access that system an additional 46 hours a week in 1998. IRS
was not able to provide us with complete data on the number of
taxpayers who took advantage of the extended hours of service.

During the 1998 filing season, IRS expanded its use of voice
messaging to increase the number of calls answered. As we reported
last year, IRS studied several areas of complicated tax law and
determined that certain call topics resulted in assistors'
spending a significantly longer time per call. 41 As a result of
the study, IRS revised its procedures in 1997 so that callers with
questions in complex tax areas were automatically connected to a
voice messaging system. For the 1998 filing season, IRS added 14
topics to the voice messaging system, increasing the number of
topics to 23. In 1998, a recording instructed callers to leave
their name, address, telephone number, and the best time for IRS
to call them back. The recording stated that an IRS representative
would attempt to return the call, at the requested time, within 3
business days. 42

As was done in 1997, IRS' examination function (Exam) supported
the customer service function by detailing staff to answer calls
on the voice messaging system. A cognizant official told us that,
in total, Exam used about 368 full- time equivalent (FTE) staff
years in supporting the customer service function in 1998. IRS
estimated that the use of Exam staff in 1998 to support customer
service resulted in about $98.7 million in forgone revenue because
these staff were not available to audit returns. We did not assess
the validity of that estimate.

IRS data show that 1.3 million calls were received by the voice
messaging system during the 1998 filing season. Of the calls
received, about 400,000 could not be returned because IRS did not
have enough information to do so (i. e., the message was not
understandable or the taxpayer hung up before providing all of the
necessary information). Of the remaining

40 IRS' standard hours of telephone assistance in 1998 were 7 a.
m. to 11 p. m., Monday through Saturday. In 1997, the standard
hours of telephone assistance were 7: 30 a. m. to 5: 30 p. m.,
Monday through Friday, with assistance on selected Saturdays
during the filing season.

41 GAO/GGD-98-33. 42 The recording during the 1997 filing season
stated that the taxpayer's call would be returned within 2
business days. An IRS official told us that the 2- day standard
was not changed for the 1998 filing season but the messaging
script was incorrectly changed to 3 days. The cost of re-
recording the message to reflect the 2- day standard was not
deemed cost- effective, especially since the standard was less
than the time frame stated in the message. Extended Hours of
Service

Expanded Use of Voice Messaging

B-279310 Page 32 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

900,000 calls, IRS returned about 735,000 (about 82 percent). IRS
officials told us that IRS does not plan to use voice messaging to
handle complex tax questions during the 1999 filing season because
it expects to have sufficient staff to handle all incoming calls.
According to the officials, those plans are subject to change if
call volumes exceed expectations.

To support the extended hours of service previously mentioned, IRS
increased staffing for telephone assistance. Additional IRS staff
were detailed to answer calls during periods of heavy call volume
so that as many calls as possible could be answered on the
taxpayer's first attempt to contact IRS. An IRS official projected
that through the end of fiscal year 1998, IRS would expend about
8,167 FTEs an increase of about 18 percent compared with the
fiscal year 1997 expenditure of about 6,892 FTEs.

IRS data show that more calls were answered on the first call
attempt during the 1998 filing season. The average number of calls
each taxpayer had to make before reaching IRS decreased to about
1. 1 from about 1.4 during the 1997 filing season. 43 Fewer repeat
callers indicates that IRS answered more calls on the first call
attempt, thereby reducing the number of times a taxpayer had to
call back. That, in turn, reduced the overall demand on IRS'
telephone system. IRS telephone data also show a significant
reduction in the number of busy signals that taxpayers experienced
when attempting to contact IRS during the 1998 filing season. IRS
reported about 4.5 million busy signals in 1998 (a busy rate of
about 9 percent) compared to about 25 million busy signals in 1997
(a busy rate of about 36 percent). 44 In comparing 1997 and 1998
telephone data, IRS had almost an 82- percent decrease in the
number of busy signals.

In another attempt to increase accessibility, IRS offered
taxpayers who were on hold waiting for an assistor for more than
10 minutes the option of leaving a message and getting a return
call from IRS. IRS used voice messaging to help reduce the call
queue and to allow more taxpayers into the telephone system for
assistance. An IRS official told us that this change gave
taxpayers an option instead of having to wait in the queue and
becoming frustrated. IRS did not have any data on the number of
callers who took advantage of this option. Although IRS does not
plan to automatically direct certain callers to a voice messaging
system in 1999, as

43 The information on average number of call attempts per taxpayer
does not include data from the form- ordering telephone line. 44
The busy signal rate is calculated by dividing the total number of
busy signals received by the total number of call attempts.
Increased Staffing

Taxpayers on Hold Given Callback Option

B-279310 Page 33 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

it did in 1998, it plans to continue to give callers who have been
on hold for more than 10 minutes the option to leave a message.

Changes to IRS' method of routing calls also may have contributed
to increased accessibility. In the past, IRS distributed call
traffic on the basis of the area code of the originating call. For
example, a call site would receive calls from taxpayers in area
codes served by a specific service center. According to IRS,
routing problems would occur when call traffic exceeded the
capacity of the incoming telephone lines at those call sites. When
call traffic surged from the area codes that the call site was
assigned to answer, it was often difficult to re- route calls to
other sites. Re- routing calls required contact with the regional
office, the National Office, and the telephone vendor, which
caused delays. Under this routing method, the number of busy
signals received by callers increased.

During the 1998 filing season, IRS revised its method of
distributing calls by changing from area code- based call routing
to a nationwide call allocation routing plan. Under that plan,
incoming call traffic is routed to all available call sites on a
percentage basis, regardless of the area code. Each call site is
staffed to answer a percentage of the total call volume, which
allows IRS to level off call traffic and balance the demand when
there is an unexpected increase. According to IRS, the change to
nationwide call allocation routing provides several benefits, such
as fewer busy signals, a reduction in the number of call attempts
per taxpayer, and increased access to the telephone system.

For the 1999 filing season, IRS plans to further upgrade its call
distribution capability with the December 1998 implementation of
an intelligent call routing system, known as the Customer Service
Call Router, which is to provide (1) instantaneous information on
the status of telephone circuits, assistor availability, abandoned
calls, and queue times and (2) a real- time capability to route
calls. This real- time routing capability should allow IRS to
better control call traffic by automatically sending calls
anywhere in the nation where call volume is lower and assistors
are available.

Besides the telephone assistance provided by IRS staff, there are
other ways that taxpayers can get information from IRS without
leaving their homes. Two of the better known methods are IRS'
World Wide Web site on the Internet and TeleTax an automated
system that provides recorded information on about 150 tax topics.
During the 1998 filing season, the Web site experienced increased
use, while the use of TeleTax declined. IRS data show a
significant level of customer dissatisfaction with TeleTax, but
there are insufficient data to explain the reasons for that
dissatisfaction. Improved Call Routing

System Taxpayers Used IRS' Web Site and TeleTax Extensively but
Expressed Dissatisfaction With TeleTax

B-279310 Page 34 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

According to IRS data, use of its Web site increased in 1998. From
January 1 to June 28, 1998, the site had recorded about 412
million hits, an 187- percent increase from the previous filing
season. A hit is recorded each time a user accesses a different
page on the Web site. Another indicator of the Web site's use is
files downloaded. Through June 28, 1998, about 29.8 million files
(tax forms, publications, regulations, and other documents) were
downloaded from the Web site, a 313- percent increase from the
same time in 1997. Another useful indicator of the value of IRS'
Web site would be the number of returns that are filed using forms
that have been downloaded from the Internet. Through various codes
imprinted on the forms, IRS tracks the source (such as tax
packages, post offices, and IRS walk- in sites) of forms that
taxpayers use to file their returns. However, because IRS uses the
same code for forms that are downloaded from the Web site as it
does for forms that taxpayers download from CD- ROMs and receive
from IRS via facsimile, it is not possible to determine to what
extent, if at all, IRS' Web site has grown as a source of forms.

We conducted a nonstatistical test to assess the ease and speed
with which forms can be downloaded from the Web site. Each weekday
from March 30 through April 15, 1998, we attempted to download tax
forms at various times of the day. In each case, we were able to
download the file quickly and easily. We recognize that an
individual's ability to download files depends on the specific
hardware and software used, and, therefore, others might have
different results from ours.

Taxpayers can also submit tax law questions via E- mail through
the Web site. Between January 1 and April 15, 1998, about 82,000
questions were submitted, up from about 44,000 questions for the
same period in 1997. IRS conducted a customer satisfaction survey
that showed that 95 percent of those responding were satisfied
with the timeliness of IRS' response to their E- mail questions.
About 75 percent of the respondents said that IRS' response
answered their questions, and almost all who responded said that
they would use the E- mail service in the future.

Taxpayers used TeleTax extensively during the 1998 filing season
but to a lesser extent than they did in 1997. IRS data show that
from January 1 to April 18, 1998, taxpayers made 37.5 million
calls to the TeleTax system versus 44.2 million calls made during
the same period in 1997 about a 15percent decrease.

IRS conducted a survey of selected taxpayers who used TeleTax by
playing a recorded message after the taxpayer had finished
listening to a TeleTax topic. The survey asked one question: Was
the taxpayer completely Use of Internet Increased

During the 1998 Filing Season

IRS Data Indicate a High Level of Taxpayer Dissatisfaction With
TeleTax

B-279310 Page 35 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

satisfied, partially satisfied, or dissatisfied with the
assistance provided by TeleTax? IRS data from this survey for the
1996 and 1997 filing seasons showed that a majority of the
surveyed users expressed some dissatisfaction with the service
(the results of the survey for 1998 were not available at the time
we completed our audit work for this report). In 1997, for
example, about 44 percent of the respondents said that they were
dissatisfied with TeleTax, and another 22 percent said that they
were only partially satisfied.

The survey did not ask any follow- up questions that could be used
to explain why taxpayers were dissatisfied. When we asked
officials how IRS uses the survey data, we were told that the data
are used to determine the level of taxpayer satisfaction.

IRS officials believe that much of the dissatisfaction with
TeleTax stems from difficulties taxpayers are experiencing in
accessing the system. In an August 1998 report on the results of
its on- line review of the 1998 filing season, IRS' Internal Audit
discussed its difficulties in trying to access TeleTax during a
test on February 23 and 24, 1998. 45 Internal Audit also discussed
problems it encountered after accessing the system, such as
difficulty hearing the automated instructions because of static.

IRS plans to improve TeleTax by centralizing the number of system
sites and upgrading equipment. According to IRS officials, TeleTax
will be moved in 2000 to new equipment that will add several
capabilities, such as the option to transfer a call to a live
assistor and some interactive scenarios that will enable callers
to get information that is specific to their circumstances.

IRS currently uses the Distributed Input System to process tax
returns and the Remittance Processing System to process tax
payments. Because both systems are old and are not Year 2000
compliant, IRS plans to replace them at its 10 service centers
with 1 integrated system, which is known as the Integrated
Submission and Remittance Processing (ISRP) System. Although the
system is integrated, each function (return processing and
remittance processing) operates independently. The Austin Service
Center tested ISRP in 1998. We monitored this test as part of our
review of the 1998 filing season because of ISRP's importance in
ensuring that IRS can meet the demands of future filing seasons.

45 On- Line Review of the 1998 Filing Season, IRS Internal Audit,
Reference No. 085408, August 7, 1998. IRS Is Moving Forward

With Its New Processing System Without a Contingency Plan for 2000

B-279310 Page 36 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

IRS officials reported the following results during the ISRP test:

 Austin did not process 100 percent of the returns that it
received during the 1998 filing season because (1) it did not
receive the number of terminals necessary to handle its return
volume until after the peak of the filing season and (2) ISRP was
not configured to process some forms that accounted for a large
volume of Austin's returns until after the filing season ended.

 Austin processed 100 percent of its remittance volume through
ISRP, but encountered many system problems that caused processing
delays and downtime. The contract contained a requirement that
ISRP be able to process 130,000 remittances in 2, 10- hour shifts.
Austin was never able to process that number of remittances
through ISRP in the allotted time.

 The contract requirement that ISRP operate at a 99- percent level
of effectiveness for 30 consecutive days was not met.

 The vendor delayed scheduled delivery of some software on more
than one occasion, and, on another occasion, IRS rejected
delivered software due to a high failure rate. The delayed and
rejected delivery of software required the vendor to add
additional deliveries that, as a result, compressed the time that
IRS had allocated to test the software.

IRS' Internal Audit reviewed development of the ISRP test and
reported its findings in January 1998. 46 Internal Audit concluded
that the ISRP project did not follow a disciplined development
process. Specifically, the remittance processing function contains
significant enhancements, such as an imaging capability and an
archiving and retrieval capability. Internal Audit found that IRS
had not developed a case to support these enhancements, nor did it
analyze whether the enhancements were needed or whether their cost
was justified.

Most of the remittance processing problems identified during the
test involved the new imaging capability. One significant problem
Austin encountered was the imaging equipment's inability to read
most of the money orders that Austin received. According to the
ISRP project manager, ISRP could not capture an image of many
money orders, which represent about 25 to 33 percent of Austin's
remittance volume, because the print on the money orders was too
light and/ or numbers showing the dollar amount were too large.
When ISRP cannot capture an image, the record is rejected to an
error file. Correction of these errors required going back to the
money order to obtain the remittance information and then

46 Review of the Initial System Development Activities of the
Integrated Submission and Remittance Processing System, IRS
Internal Audit, Reference No. 082204, January 30, 1998. Results of
ISRP Test

B-279310 Page 37 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

manually entering that information. The amount of manual effort
required decreased the system's productivity and delayed the
processing of remittances.

Another concern with ISRP's remittance processing function is the
equipment's need for specialized care that requires trained
operators to spend time each day to keep the equipment cleaned and
operating properly. Adjustments must be made to the mechanism that
transports the documents through the equipment to keep them
feeding through and to ensure that the scanner is reading the
correct area of the document.

The problems previously discussed led IRS to revise Austin's
original contingency plan for processing remittances. Austin's
original contingency plan was to remove the old processing
equipment while leaving cables in place to facilitate quick
reinstallation of that equipment if it was needed. As a result of
the test, IRS has decided that service centers should keep enough
of the old equipment fully installed for 1999 so that processing
can be switched to the old systems if there are problems with
ISRP. However, four centers do not have enough space to house both
the old equipment and ISRP.

In part because of the new contingency plan and its impact on
space needs, IRS revised its time frame for implementing ISRP.
Installation of both the return processing and remittance
processing functions of ISRP was to be completed at all service
centers by November 1998. In May 1998, IRS revised the
implementation schedule for the remittance processing function to
delay until August 1999 installation at the four centers that do
not have enough space for both the old and new equipment. Until
August 1999, the four centers plan to continue using the old
remittance processing equipment.

Although the decision to delay implementation of the remittance
processing function in four centers may reduce the risk of serious
processing problems in 1999, we are concerned about the potential
impact in 2000. The four service centers scheduled to receive the
remittance processing function in August 1999 were among the top
five centers in the volume of remittances processed during the
peak of the 1998 filing season (the center with the largest volume
Ogden is one of the centers where the remittance processing
function was installed in 1998). We are concerned that (1) these
four centers will receive the remittance processing function so
late in 1999 that they will have no experience processing the
large volume of remittances that come in at the peak of the filing
season and (2) August 1999 is close to the Year 2000 deadline when
IRS Revised Its ISRP

Contingency Plan for 1999 but Has No Contingency Plan for 2000

B-279310 Page 38 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

the old systems will no longer function. IRS has not developed a
contingency plan for the processing of remittances beyond the 1999
filing season.

Using various IRS indicators as criteria, the 1998 filing season
was generally a success. For example, there were significant gains
in the ability of taxpayers to reach IRS over the telephone and in
the use of electronic filing. Also, taxpayer use of recent
enhancements, such as the ability to obtain information and
download forms from IRS' Web site, increased significantly in
1998. Available data also indicate that thousands of taxpayers
took advantage of the increased EIC- targeted assistance offered
by IRS in 1998, and that IRS' enhanced compliance efforts
apparently contributed to fewer SSN- and EIC- related errors in
1998. IRS had to delay processing some returns in 1998 until it
had made the programming changes needed to implement new
legislation on capital gains, but we saw no evidence that those
delays caused undue hardships to taxpayers. It appears, however,
that IRS' procedure for processing returns that were filed without
a required Schedule D increased taxpayer burden unnecessarily. By
suspending the processing of returns until the taxpayers submitted
a Schedule D, IRS caused those taxpayers to complete a complicated
and time- consuming form even when the capital gain rates would
have little, if any, effect on their tax liability. IRS has
changed its procedure for 1999, which, if effectively implemented,
should reduce the burden for some taxpayers.

Although IRS' data indicate a generally successful filing season,
there is one important filing season activity walk- in assistance
for which IRS lacks meaningful national performance data. For
example, although IRS had guidance requiring regional reviews of
the service being provided by walk- in sites during the 1998
filing season, the guidance was not specific as to what should be
reviewed, and there was no requirement that review results be
reported to the National Office. Also, IRS generally did not begin
surveying walk- in site customers until mid- March 2- 1/ 2 months
after the filing season began. Without meaningful nationwide
performance data, IRS cannot determine if the walk- in program is
meeting its objectives and goals.

Taxpayers' access to IRS' Web site or hits increased
significantly, but because tax forms downloaded in 1998 were coded
the same as forms taxpayers obtained from CD- ROMs or by
facsimile, IRS cannot evaluate the actual usefulness of its site
in providing taxpayers with the forms they need to file their
returns. A separate code would enable IRS to determine Conclusions

B-279310 Page 39 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

how many returns are being filed using forms downloaded from the
Web site.

IRS is to be commended for its strides in improving telephone
accessibility. There is one aspect of that service, however, that
needs particular attention the high level of taxpayer
dissatisfaction with TeleTax. IRS recognizes that there are
problems with TeleTax and improvements are planned. Neither we nor
IRS know whether those changes are likely to resolve all of the
critical problems with TeleTax, because the TeleTax customer
satisfaction survey provided no data on why taxpayers were
dissatisfied.

The gain in electronic filing is also commendable. But, it is
important to note that about 80 percent of the individual income
tax returns filed in 1998 were still filed on paper. IRS
recognizes the need to continue working to eliminate barriers to
electronic filing and plans to conduct various tests in 1999
directed at one of the more significant barriers the fact that
persons using traditional electronic filing still have to send
paper to IRS.

One change that IRS has resisted making would involve including
the tax table in the TeleFile tax package. We realize that the tax
table is not needed to use TeleFile and that its inclusion would
add to the program's cost by increasing the size of the TeleFile
tax package. However, 72 percent of the respondents to IRS' most
recent nonusers survey said that they would be encouraged to use
TeleFile if they were able to determine their federal tax before
calling TeleFile. We believe that IRS could structure a test that
would minimize additional cost by including the tax table in only
some TeleFile packages and comparing the filing patterns of
persons who received those packages with those who received
packages without the tax table. That test should help IRS
determine whether the inclusion of the tax table in the TeleFile
package would increase the use of TeleFile and whether the extent
of that increase would justify the additional cost of including
the tax table.

Although there is some evidence that IRS' growing emphasis on EIC
noncompliance has had a positive effect, there is much that is
still unknown. At the time we completed our audit work, data were
not available to assess the effect of the notices IRS mailed in
December 1997 or of the various projects IRS had initiated to
target specific areas of EIC noncompliance. It was also too early
to assess the impact of provisions in TRA97 that authorize IRS to
deny EIC claims in subsequent years from persons who were found to
have filed fraudulent or negligent EIC claims, because 1999 is the
first year that IRS can deny subsequent EIC claims

B-279310 Page 40 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

under the provisions. Even then, the full impact of those
provisions could not be determined because there is no information
on the extent to which knowledge of the provisions deterred
persons from filing improper EIC claims. In that regard, we
recommended in a previous report that IRS enhance the potential
deterrent effect by better publicizing the provisions in the Form
1040 tax package. IRS has agreed to change the Schedule EIC
instructions in the package but does not plan to add anything to
the schedule itself. We believe that a brief, but prominent,
cautionary statement added to the Schedule EIC would alert
taxpayers to important information in the instructions that they
should read before filing their returns. There is also reason to
believe that EIC criteria may not have been applied consistently
at all 10 service centers because of a lack of controls to ensure
that field locations receive timely notification of important
procedural changes.

Since our report on the 1997 filing season and in response to a
recommendation in that report, FMS completed a study that supports
the assumption that having lockboxes process Form 1040 tax
payments saves the government money, although not as much as
previously assumed. In light of FMS' study results, we are not
suggesting that IRS stop using lockboxes to process those
payments. However, because IRS has not developed the kind of
definitive data that we recommended in our report on the 1997
filing season, we believe that IRS still lacks the necessary
evidence to determine whether the additional burden and taxpayer
costs that may be caused by having taxpayers separate their
returns from their payments outweigh the additional governmental
cost associated with having the banks handle those returns.

Considering the various problems encountered by the Austin Service
Center in processing remittances during the test of ISRP in 1998,
rollout of the system to other locations could have significant
implications for upcoming filing seasons. IRS has developed a
contingency plan for the 1999 filing season, when the other nine
service centers are to begin processing returns through ISRP and
five of the nine are to begin using ISRP for remittances. However,
IRS has no contingency plan for the 2000 filing season. We believe
that IRS' decision to wait until August 1999 to install ISRP's
remittance processing function in the other four centers adds an
element of risk to the 2000 filing season that warrants further
contingency planning.

B-279310 Page 41 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

We recommend that the Commissioner of Internal Revenue direct the
appropriate officials to take the following steps:

 Conduct a test to determine (1) if adding the tax table to the
TeleFile package would increase TeleFile use and, if so, (2)
whether the increased use would justify the additional cost of
including the tax table in the package. The cost of such a test
can be minimized if IRS includes the tax table only in the
TeleFile packages sent to a representative sample of taxpayers and
then compares their filing patterns to the filing patterns of
taxpayers who received TeleFile packages without the tax table. If
IRS determines, after designing the test, that its cost would be
prohibitive, IRS should consider other options for dealing with
the issue raised by the TeleFile nonusers.

 Develop controls and standardized procedures to ensure that field
locations are notified of procedural changes and implement the
changes at the same time.

 In conjunction with IRS' decision to revise the instructions for
Schedule EIC to advise taxpayers of the 2- year and 10- year
sanctions, add a brief, but prominent, cautionary statement to the
Schedule EIC alerting taxpayers that they should read important
information in the instructions before filing their returns. We
realize that it is too late to make such a change to the Schedule
EIC for tax year 1998 and that the first opportunity to make this
change will be for tax year 1999.

 To provide a better measure of the extent to which taxpayers are
using the Web site, use a separate code to identify forms that
have been downloaded from the site.

 Expand future TeleTax customer satisfaction surveys to obtain
information on why respondents are dissatisfied.

 Develop a contingency plan for ISRP that provides for the
possibility of a systemwide failure of the remittance processing
function past 1999.

We are not making a specific recommendation relating to the
measurement of walk- in services because we previously made a
relevant recommendation in our February 1998 report on measuring
customer service. 47 In that report, we recommended that the
Commissioner direct the appropriate officials to develop
performance indicators that cover the full range of IRS' customer
service programs.

47 GAO/GGD-98-59. Recommendations to

the Commissioner of Internal Revenue

B-279310 Page 42 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

We obtained IRS' written comments on a draft of this report in a
December 15, 1998, letter from IRS' Deputy Commissioner Operations
(see app. III). We also met with IRS' Chief Operations Officer and
other IRS officials who are responsible for the various programs
we reviewed on December 4, 1998, to discuss their comments. IRS
said that, except for our first and last recommendations regarding
TeleFile and ISRP, it generally agreed with the findings and
recommendations in our draft report.

IRS did not agree with our recommendation that it conduct a test
to determine if adding the tax table to the TeleFile package might
increase the use of TeleFile and, if so, whether the increased use
would justify the additional cost associated with including the
tax table. According to IRS, while a past survey indicated that
some taxpayers might have used TeleFile if they could have
determined the tax themselves, the data does not warrant the
addition of the tax tables, instructions, and worksheet at a cost
of almost $1.4 million. IRS also said that it had not received
similar comments from taxpayers in recent years.

IRS' reference to some taxpayers in discussing the nonusers survey
understates the results of the survey, which showed that a large
majority (72 percent) of the nonusers said that they would have
been encouraged to use TeleFile if they had been able to determine
their federal tax beforehand. Also, IRS' statement that it has not
received similar comments in recent years is confusing because (1)
the 1997 nonusers survey was recent and (2) nonusers did not have
an opportunity to provide similar comments in 1998 because IRS did
not do a nonusers survey that year.

Although we did not verify IRS' estimate of $1.4 million in
additional costs, we recognize that inclusion of the tax table and
related materials in all TeleFile packages would be costly. That
is why we recommended a test. In our opinion, such a test is
needed to determine whether inclusion of the tax table would
increase TeleFile use enough to warrant the additional cost.

IRS also said that the addition of the tax table and other
material that would be needed for taxpayers to determine their tax
liability would increase the complexity of the TeleFile tax
package. Although adding the tax table and other material to the
TeleFile package would increase the package's size, we see no
reason why it would increase TeleFile's complexity. Inclusion of
the tax table and other material would give those taxpayers who
want to calculate their tax before calling TeleFile the ability to
do so; other taxpayers could simply ignore that part of the
package. Agency Comments and

Our Evaluation

B-279310 Page 43 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

While we continue to believe that a test is needed, we are
sensitive to the potential costs involved and do not want to
suggest that a test should be done no matter how costly. At the
same time, we believe that efforts to substantially increase the
use of TeleFile require that IRS heed the results of the nonusers
survey. If, after designing a potential test, IRS determines that
the cost of doing the test would be prohibitive, it should
consider other options for dealing with the issue raised by the
TeleFile nonusers. One option might be to devise a communication/
marketing strategy directed at allaying any concerns about
TeleFile that would cause taxpayers to want to determine their tax
liability themselves before using TeleFile. To acknowledge the
possibility that the cost of a test might be prohibitive and that
other options might be more feasible, we have revised our
recommendation in finalizing this report.

Regarding our recommendation that IRS develop controls and
standardized procedures to ensure that field locations are
notified of procedural changes and implement the changes at the
same time, IRS said that the system used to document and
communicate program and procedural changes does not have a feature
to notify the originator that an office has accessed the
information. IRS said that it will pursue adding such a feature as
a future enhancement to the system. IRS also said that as part of
its readiness for the 1999 filing season, it will reemphasize to
service center directors the importance of ensuring that their
field locations expeditiously access program and procedural
changes and disseminate and implement all required actions in a
timely manner. These actions, if effectively implemented, will
meet the intent of our recommendation.

IRS agreed with our recommendation that it add a cautionary
statement to the Schedule EIC alerting taxpayers to potential
sanctions. IRS said that it would revise the 1999 Schedule EIC to
(1) alert taxpayers that they should read the penalties that apply
if they improperly claim the EIC and (2) refer taxpayers to the
EIC instructions for that information.

IRS also agreed with our recommendation that it use a separate
code to identify forms that have been downloaded from IRS'
Internet Web site. In our December 4 meeting, IRS officials said
that IRS had implemented separate codes for reporting the source
of electronically produced forms (i. e., Internet, CD- ROM, and
facsimile) in 1996, but had experienced problems with certain off-
the- shelf software when producing the CD- ROM for the tax year
1997 forms that resulted in its inability to designate separate
codes in 1998. According to the officials, IRS has resolved the

B-279310 Page 44 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

source code issue and will have separate source codes for
reporting tax year 1998 forms generated by CD- ROM, the Internet,
and facsimile.

Regarding our recommendation that IRS expand future TeleTax
customer satisfaction surveys, IRS said that the former survey has
been discontinued and that the current TeleTax system is scheduled
to be replaced beginning in 2000. In conjunction with that
replacement, according to IRS officials, IRS intends to reinstate
an improved customer satisfaction survey that will include
additional questions and the ability to capture taxpayer comments.
IRS' decision to discontinue the survey until TeleTax is replaced
and an improved survey can be put in place was consistent with the
intent of our draft recommendation, so we revised our
recommendation to indicate these plans.

IRS disagreed with the last recommendation in our draft report,
which called for IRS to develop a contingency plan for ISRP that
provides for the possibility of extended downtime of the
remittance processing function past 1999. IRS said that it had
developed a contingency plan to be used after 1999. In explaining
that plan, IRS said that (1) normal disaster recovery procedures
will be in place in case of an extended downtime of remittance
processing equipment and (2) it will have in place a system to
direct payments received to lockbox facilities as needed. We did
not receive a copy of IRS' normal disaster recovery procedures in
time to analyze them. However, cognizant IRS officials told us
that the procedures are service- center specific and focus on the
transshipment of work from one center to another, not the movement
of work from a service center to a lockbox. They also told us that
the procedures did not cover systemwide failures, when
transshipment between service centers would be of no value.
Regarding the potential use of lockboxes to handle the increased
remittance processing workload if the service centers cannot, a
cognizant FMS official told us that such a change would require
negotiations with the lockboxes. According to the official, no
such negotiations had begun as of December 4, 1998.

After considering IRS' comments on our last recommendation, we
have retained our recommendation but changed its wording to
clarify that our concern centers on the potential for a systemwide
failure.

We are sending copies of this report to the Subcommittee's Ranking
Minority Member, the Chairmen and Ranking Minority Members of the
House Committee on Ways and Means and the Senate Committee on
Finance, various other congressional committees, the Secretary of
the Treasury, the Commissioner of Internal Revenue, the Director
of the Office

B-279310 Page 45 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

of Management and Budget, and other interested parties. We will
also make copies available to others on request.

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

Sincerely yours, James R. White Director, Tax Policy and

Administration Issues

Page 46

Contents 1 Letter 48 Appendix I Toll- Free Telephone Accessibility
Test

53 Appendix II IRS Workload Indicators

54 Appendix III Comments From the Internal Revenue Service

59 Appendix IV Major Contributors to This Report

Table 1: IRS' Performance Goals and Related Accomplishments for
the 1997 and 1998 Filing Seasons

7 Table 2: Number of Individual Income Tax Returns

Received, by Filing Type 14

Table 3: Accessibility of IRS' Telephone Assistance 47 Tables

Figure I. 1: Results of GAO's Telephone Accessibility Test 49
Figure I. 2: Level of Access and Composition 50 Figure I. 3: Level
of Service and Composition 51 Figure I. 4: Message Callback Rate
52 Figures

Contents Page 47 Abbreviations

EIC Earned Income Credit FMS Financial Management Service FTE
full- time equivalent IRS Internal Revenue Service ISRP Integrated
Submission and Remittance Processing SSN Social Security number
TRA97 The Taxpayer Relief Act of 1997

Appendix I Toll- Free Telephone Accessibility Test

Page 48 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

To independently test the accessibility of IRS' toll- free
telephone assistance, we conducted a nonprojectable test by making
up to six calls per day to IRS' tax law assistance telephone
number (1- 800- TAX- 1040) from four metropolitan areas Atlanta,
GA; Kansas City, KS; San Francisco, CA; and Washington, D. C. We
made the calls during the following two periods in 1998: from
February 1 through February 26 and from March 30 through April 15.
To avoid possible bias, each call attempt was randomly spaced
throughout the day from 7 a. m. to 11 p. m. Each attempt to
contact IRS consisted of up to five calls that were spaced 1
minute apart. If we received a busy signal, we hung up, waited
approximately 1 minute, then redialed. If after four redials (five
calls in total) we had not reached IRS, we considered the attempt
unsuccessful. In conducting our test, we did not ask questions of
the assistors because it was not our intent to assess the accuracy
of their answers.

Once in the telephone system, we were either routed to an assistor
or to the messaging system, depending on the topic selected from
the menu. When routed to an assistor, we (1) abandoned the call if
we were on hold for 7 minutes and (2) did not dial again. We
considered abandoned calls as successful attempts in computing
level of access because we were able to access the telephone
system. However, we considered those calls unsuccessful in
computing level of service because we did not make contact with an
assistor. For calls routed to the messaging system, we left a name
and telephone number that had message recording capabilities. If
we received a call from IRS within 3 business days, we considered
it a successful attempt (call answered) and included it in our
calculation of the message callback rate.

For our test, we measured access to IRS' telephone system (level
of access), the rate at which calls were answered (level of
service), and the callback rate from the messaging system. See
figure I. 1 for the overall results of our test.

Appendix I Toll- Free Telephone Accessibility Test

Page 49 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Source: GAO test.

We made a total of 786 call attempts. Of those 786 attempts, 635
resulted in access to the system (an 81- percent level of access).
Of the 635 accesses, we reached an assistor 477 times, left a
message and received a return call from IRS 112 times, and
abandoned the call after being on hold for 7 minutes 46 times. Of
the 151 unsuccessful attempts, 131 resulted in busy signals and 20
were calls to the messaging system that were not returned by IRS.
These results are depicted in figure I. 2.

Figure I. 1: Results of GAO's Telephone Accessibility Test

Level of Access

Appendix I Toll- Free Telephone Accessibility Test

Page 50 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Source: GAO test.

As measured by IRS, the only difference between level of access
and level of service is that the former is computed by including
abandoned calls in the number of calls that gained access (i. e.,
the callers had gained access to IRS' system but subsequently
decided, for unknown reasons, to hang up before an assistor came
on the line), while the latter is computed after deleting the
abandoned calls (i. e., although those callers gained access to
the system, they were not served). Thus, to compute the level of
service from our test, we deleted the 46 abandoned calls that were
included in our computation of level of access. As shown in figure
I. 3, of our 786 call attempts, we either reached an assistor or
got a return call from IRS 589 times (a 75- percent level of
service).

Figure I. 2: Level of Access and Composition

Level of Service

Appendix I Toll- Free Telephone Accessibility Test

Page 51 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Source: GAO test.

Of our 786 call attempts, 132 were routed to IRS' messaging
system. IRS told callers leaving messages that an IRS
representative would attempt to return their calls within 3
business days. Of our 132 call attempts that went to the messaging
system, 112 resulted in a return call from IRS within 3 business
days (a callback rate of 85 percent). That result is depicted in
figure I. 4.

Figure I. 3: Level of Service and Composition

Message Callback Rate

Appendix I Toll- Free Telephone Accessibility Test

Page 52 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Source: GAO test.

Figure I. 4: Message Callback Rate

Appendix II IRS Workload Indicators

Page 53 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Indicator As of date 1997 (thousands) 1998 (thousands) Percentage
change a

Individual income tax returns received Alternative filing methods:

Traditional electronic 10/ 30 14,457 18,639 28.93 Form 1040 PC 10/
30 8,427 7,533 -10.61 TeleFile 10/ 30 4,694 5,963 27.03 Total, all
methods 10/ 30 120,941 123,055 1. 75 Refunds sent Number 10/ 30
81,865 82,998 1. 38 Amount 10/ 30 $105,664,000 $112,308,000 6.29
Number of direct deposits 10/ 30 16,469 19,225 16.73 Amount of
direct deposits 10/ 30 $29,051,000 $34,543,000 18.90 Extension of
time to file 5/ 15 6,548 6,976 6.54 Internet use Hits 6/ 28
143,662 411,758 186.62 Files downloaded 6/ 28 7,222 29,826 313.00
Tax forms faxed to taxpayers 6/ 28 671 1,003 49.53 Receipts b
Total number of receipts 8/ 31 202,855 208,053 2. 56 Total amount
of receipts 8/ 31 $1,441,187,407 $1,572,198,961 9. 09 Total
TeleTax calls 4/ 18 44,170 37,463 -15.19 Tax law calls 4/ 18 6,
765 7,728 14.22 Refund calls 4/ 18 37,405 29,735 -20.51 Total
toll- free telephone calls c 4/ 18 70,204 48,485 -30.94 Calls
answered 4/ 18 36,910 35,730 -3.20 Busy signals 4/ 18 24,953 4,
535 -81.83 Abandons 4/ 18 8,341 8,220 -1.45 Taxpayers assisted at
walk- in sites 4/ 25 5, 988 6,173 d 3.10 EIC Number of recipients
8/ 29 19,032 19,393 1. 90 Total amount of EIC 8/ 29 $27,811,508
$29,406,298 5. 73

a Numbers may not compute to this percentage due to rounding. b
Total receipts include remittances from individuals and
businesses. c Toll- free telephone calls include calls to the
following phone lines: (1) tax law assistance, (2) EIC and/ or
refund inquiry, (3) account inquiry, (4) form ordering, and (5)
fraud hotline. d Included in this number are 82,000 taxpayers
served at walk- in sites on Saturdays from March 7

through April 11, 1998. There were about 6.1 million taxpayers
served during the weekdays in 1998. Source: IRS data.

Appendix III Comments From the Internal Revenue Service

Page 54 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Appendix III Comments From the Internal Revenue Service

Page 55 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Appendix III Comments From the Internal Revenue Service

Page 56 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Appendix III Comments From the Internal Revenue Service

Page 57 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Appendix III Comments From the Internal Revenue Service

Page 58 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Appendix IV Major Contributors to This Report

Page 59 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

David J. Attianese, Assistant Director, Tax Policy and
Administration Issues M. Grace Haskins, Senior Evaluator John
Lesser, Senior Evaluator Monika R. Gomez, Evaluator

Royce L. Baker, Issue Area Manager Doris J. Hynes, Evaluator- in-
Charge Marvin G. McGill, Senior Evaluator Rose M. Dorlac,
Evaluator Bradley L. Terry, Evaluator

Jyoti Gupta, Evaluator Susan S. Mak, Evaluator General Government

Division, Washington, D. C.

Kansas City Field Office

Atlanta Field Office San Francisco Field Office

Page 60 GAO/GGD-99-21 IRS' 1998 Tax Filing Season

Ordering Information The first copy of each GAO report and
testimony is free. Additional copies are $2 each. Orders should be
sent to the following address, accompanied by a check or money
order made out to the Superintendent of Documents, when necessary.
VISA and MasterCard credit cards are accepted, also. Orders for
100 or more copies to be mailed to a single address are discounted
25 percent.

Order by mail: U. S. General Accounting Office P. O. Box 37050
Washington, DC 20013

or visit: Room 1100 700 4 th St. NW (corner of 4 th and G Sts. NW)
U. S. General Accounting Office Washington, DC

Orders may also be placed by calling (202) 512- 6000 or by using
fax number (301) 258- 4066, or TDD (301) 413- 0006.

Each day, GAO issues a list of newly available reports and
testimony. To receive facsimile copies of the daily list or any
list from the past 30 days, please call (202) 512- 6000 using a
touch- tone phone. A recorded menu will provide information on how
to obtain these lists.

For information on how to access GAO reports on the INTERNET, send
e- mail message with info in the body to:

info@ www. gao. gov or visit GAO's World Wide Web Home Page at:
http:// www. gao. gov

United States General Accounting Office Washington, D. C. 20548-
0001

Official Business Penalty for Private Use $300

Address Correction Requested Bulk Rate

Postage & Fees Paid GAO Permit No. G100

(268831)

*** End of document. ***