[House Hearing, 108 Congress]
[From the U.S. Government Printing Office]




 
             IRS EFFORTS TO MODERNIZE ITS COMPUTER SYSTEMS

=======================================================================

                                HEARING

                               before the

                       SUBCOMMITTEE ON OVERSIGHT

                                 of the

                      COMMITTEE ON WAYS AND MEANS
                     U.S. HOUSE OF REPRESENTATIVES

                      ONE HUNDRED EIGHTH CONGRESS

                             SECOND SESSION

                               __________

                           FEBRUARY 12, 2004

                               __________

                           Serial No. 108-34

                               __________

         Printed for the use of the Committee on Ways and Means


____________________________________________________________________________
For Sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpo.gov  Phone: toll free (866) 512-1800; (202) 512�091800  
Fax: (202) 512�092250 Mail: Stop SSOP, Washington, DC 20402�090001

                      COMMITTEE ON WAYS AND MEANS

                   BILL THOMAS, California, Chairman

PHILIP M. CRANE, Illinois            CHARLES B. RANGEL, New York
E. CLAY SHAW, JR., Florida           FORTNEY PETE STARK, California
NANCY L. JOHNSON, Connecticut        ROBERT T. MATSUI, California
AMO HOUGHTON, New York               SANDER M. LEVIN, Michigan
WALLY HERGER, California             BENJAMIN L. CARDIN, Maryland
JIM MCCRERY, Louisiana               JIM MCDERMOTT, Washington
DAVE CAMP, Michigan                  GERALD D. KLECZKA, Wisconsin
JIM RAMSTAD, Minnesota               JOHN LEWIS, Georgia
JIM NUSSLE, Iowa                     RICHARD E. NEAL, Massachusetts
SAM JOHNSON, Texas                   MICHAEL R. MCNULTY, New York
JENNIFER DUNN, Washington            WILLIAM J. JEFFERSON, Louisiana
MAC COLLINS, Georgia                 JOHN S. TANNER, Tennessee
ROB PORTMAN, Ohio                    XAVIER BECERRA, California
PHIL ENGLISH, Pennsylvania           LLOYD DOGGETT, Texas
J.D. HAYWORTH, Arizona               EARL POMEROY, North Dakota
JERRY WELLER, Illinois               MAX SANDLIN, Texas
KENNY C. HULSHOF, Missouri           STEPHANIE TUBBS JONES, Ohio
SCOTT MCINNIS, Colorado
RON LEWIS, Kentucky
MARK FOLEY, Florida
KEVIN BRADY, Texas
PAUL RYAN, Wisconsin
ERIC CANTOR, Virginia

                    Allison H. Giles, Chief of Staff

                  Janice Mays, Minority Chief Counsel

                                 ______

                       SUBCOMMITTEE ON OVERSIGHT

                    AMO HOUGHTON, New York, Chairman

ROB PORTMAN, Ohio                    EARL POMEROY, North Dakota
JERRY WELLER, Illinois               GERALD D. KLECZKA, Wisconsin
SCOTT MCINNIS, Colorado              MICHAEL R. MCNULTY, New York
MARK FOLEY, Florida                  JOHN S. TANNER, Tennessee
SAM JOHNSON, Texas                   MAX SANDLIN, Texas
PAUL RYAN, Wisconsin
ERIC CANTOR, Virginia

Pursuant to clause 2(e)(4) of Rule XI of the Rules of the House, public 
hearing records of the Committee on Ways and Means are also published 
in electronic form. The printed hearing record remains the official 
version. Because electronic submissions are used to prepare both 
printed and electronic versions of the hearing record, the process of 
converting between various electronic formats may introduce 
unintentional errors or omissions. Such occurrences are inherent in the 
current publication process and should diminish as the process is 
further refined.


                            C O N T E N T S

                               __________

                                                                   Page

Advisories announcing the hearing................................     2

                               WITNESSES

Internal Revenue Service, Hon. Mark W. Everson, Commissioner.....     9
Internal Revenue Service Oversight Board, Hon. Larry Levitan, 
  Member.........................................................    31
U.S. General Accounting Office, Robert F. Dacey, Director, 
  Information Security Issues....................................    42

                                 ______

Computer Sciences Corporation, Paul Cofoni.......................    51
Software Engineering Institute, Carnegie Mellon University, M. 
  Steven Palmquist...............................................    37


             IRS EFFORTS TO MODERNIZE ITS COMPUTER SYSTEMS

                              ----------                              


                      THURSDAY, FEBRUARY 12, 2004

             U.S. House of Representatives,
                       Committee on Ways and Means,
                                 Subcommittee on Oversight,
                                                    Washington, DC.

    The Subcommittee met, pursuant to notice, at 9:08 a.m., in 
room 1100, Longworth House Office Building, Hon. Amo Houghton 
(Chairman of the Subcommittee) presiding.
    [The advisory, the postponing advisory, and the 
rescheduling advisory announcing the hearing follow:]

ADVISORY FROM THE COMMITTEE ON WAYS AND MEANS

                       SUBCOMMITTEE ON OVERSIGHT

                                                CONTACT: (202) 225-7601
FOR IMMEDIATE RELEASE
October 28, 2003
OV-6

                     Houghton Announces Hearing on

             IRS Efforts to Modernize its Computer Systems

    Congressman Amo Houghton (R-NY), Chairman, Subcommittee on 
Oversight of the Committee on Ways and Means, today announced that the 
Subcommittee will hold a hearing to evaluate the Internal Revenue 
Service (IRS) efforts to modernize its computer systems and hear the 
outcome of recent independent reviews of the IRS Business Systems 
Modernization (BSM) program requested by IRS Commissioner Mark Everson. 
The hearing will take place on Tuesday, November 4, 2003, in the main 
Committee hearing room, 1100 Longworth House Office Building, beginning 
at 3:00 p.m.
      
    In view of the limited time available to hear witnesses, oral 
testimony at this hearing will be from invited witnesses only. 
Witnesses will include the Honorable Mark Everson, Commissioner of the 
Internal Revenue Service (IRS), a representative of Computer Sciences 
Corporation, and Larry Levitan, Member, IRS Oversight Board.
      

BACKGROUND:

      
    United States taxpayers pay more than $2 trillion in taxes each 
year to the IRS. The IRS's receipt of tax revenue is dependent on a set 
of computer systems that has evolved over the past 35 years. Nearly all 
IRS employees depend on these computer systems to do their daily jobs, 
including more than 70,000 who use these systems to deliver direct 
service to taxpayers. The IRS network of computer systems is comprised 
of over 100,000 individual computers, 2,779 vendor supplied software 
products, and more than 50 million lines of IRS-maintained computer 
code.
      
    The existing IRS computer systems architecture became operational 
in 1967. Although significantly upgraded since then, the basic 
architecture and processing systems date from the 1960s. Since the 
1970s, the IRS has made numerous attempts at technological 
modernization. Success of past efforts was limited, due in part to the 
IRS's lack of a coordinated, unified plan and approach for fitting all 
aspects of modernization together.
      
    The current computer systems used by the IRS inhibit its ability to 
effectively carry out the mission of administering the Nation's tax 
laws. Maintaining these systems is a major undertaking and gets more 
difficult every year as older systems replacement parts and programmers 
are harder by which to come. Here are a few examples of problems caused 
by the outdated systems:
      
      American taxpayers expect a level of service from the IRS 
that rivals the private sector, but IRS systems do not provide the 
necessary functionality.
      IRS employees must work with taxpayer data that is not 
timely, resulting in frustration for both taxpayers and employees.
      IRS has trouble accounting for funds and maintaining the 
security of critical data and systems.
      
    Addressing these crucial needs, managing the inherent risks of 
modernization, and delivering the level of service taxpayers expect are 
all goals of the IRS's BSM.
      
    Prior efforts to modernize the IRS's computer systems in the mid-
1990s were not successful. These failures prompted the IRS to enter 
into an innovative contract with Computer Sciences Corporation (CSC) in 
1998 to launch the BSM program. Under the contract, CSC assists the IRS 
to design new systems and helps to identify contractors to perform 
software development and other tasks. The CSC is referred to as the 
prime contractor, under the principal BSM contract.
      
    Since mid-1999, more than $1.3 billion has been appropriated for 
BSM, including $391 million for FY 2002 and $366 million for FY 2003. 
The House and Senate have now approved an additional $429 million for 
FY 2004. The total cost of BSM is expected to be in the range of $8 
billion.
      
    Soon after his appointment, Commissioner Everson requested an 
independent review of a critical component of the BSM program to be 
conducted by Software Engineering Institute (SEI) of Carnegie Mellon 
University. The study focuses on the history of the Customer Account 
Data Engine (CADE) project and the feasibility of future plans with 
respect to CADE. The SEI has completed its interim review, and a 
witness from SEI will appear before the Subcommittee to present its 
findings and answer questions. The CSC has also commissioned Bain and 
Company to study CADE and other aspects of BSM and they will appear and 
be prepared to discuss Bain's findings. Finally, the IRS has conducted 
an internal review of the root causes of schedule delays and cost 
increases in BSM projects. The Commissioner is expected to discuss the 
results of all of the aforementioned studies and to announce changes 
that he plans to implement.
      
    In announcing the hearing, Chairman Houghton said, ``The IRS must 
bring its systems into the 21st century to provide the high and 
efficient level of service that our taxpayers expect. We saw this back 
at our 1998 hearing, and it's critical for us to see that this program 
is on track.''
      

FOCUS OF THE HEARING:

      
    The hearing will focus on the IRS efforts to modernize its computer 
systems and on independent reviews ordered by IRS Commissioner Mark 
Everson to assess the IRS's BSM program.
      

DETAILS FOR SUBMISSION OF WRITTEN COMMENTS:

      
    Please Note: Due to the change in House mail policy, any person or 
organization wishing to submit a written statement for the printed 
record of the hearing should send it electronically to 
[email protected], along with a fax copy to 
(202) 225-2610, by the close of business Tuesday, November 18, 2003. 
Those filing written statements who wish to have their statements 
distributed to the press and interested public at the hearing should 
deliver their 200 copies to the Subcommittee on Oversight in room 1136 
Longworth House Office Building, in an open and searchable package 48 
hours before the hearing. The U.S. Capitol Police will refuse sealed-
packaged deliveries to all House Office Buildings.
      

FORMATTING REQUIREMENTS:

      
    Each statement presented for printing to the Committee by a 
witness, any written statement or exhibit submitted for the printed 
record or any written comments in response to a request for written 
comments must conform to the guidelines listed below. Any statement or 
exhibit not in compliance with these guidelines will not be printed, 
but will be maintained in the Committee files for review and use by the 
Committee.
      
    1. Due to the change in House mail policy, all statements and any 
accompanying exhibits for printing must be submitted electronically to 
[email protected], along with a fax copy to 
202/225-2610, in WordPerfect or MS Word format and MUST NOT exceed a 
total of 10 pages including attachments. Witnesses are advised that the 
Committee will rely on electronic submissions for printing the official 
hearing record.
      
    2. Copies of whole documents submitted as exhibit material will not 
be accepted for printing. Instead, exhibit material should be 
referenced and quoted or paraphrased. All exhibit material not meeting 
these specifications will be maintained in the Committee files for 
review and use by the Committee.
      
    3. Any statements must include a list of all clients, persons, or 
organizations on whose behalf the witness appears. A supplemental sheet 
must accompany each statement listing the name, company, address, 
telephone and fax numbers of each witness.
      
    The Committee seeks to make its facilities accessible to persons 
with disabilities. If you are in need of special accommodations, please 
call 202-225-1721 or 202-226-3411 TTD/TTY in advance of the event (four 
business days notice is requested). Questions with regard to special 
accommodation needs in general (including availability of Committee 
materials in alternative formats) may be directed to the Committee as 
noted above.

                                 

                * * * NOTICE--HEARING POSTPONEMENT * * *

ADVISORY FROM THE COMMITTEE ON WAYS AND MEANS

                       SUBCOMMITTEE ON OVERSIGHT

                                                CONTACT: (202) 225-7601
FOR IMMEDIATE RELEASE
October 30, 2003
OV-6-Revised

                Postponement of Subcommittee Hearing on

             IRS Efforts to Modernize its Computer Systems

    Congressman Amo Houghton (R-NY), Chairman of the Subcommittee on 
Oversight of the Committee on Ways and Means, today announced that the 
Subcommittee hearing on IRS efforts to modernize its computer systems, 
previously scheduled for Wednesday, November 4, 2003, at 3:00 p.m., in 
the main Committee hearing room, 1100 Longworth House Office Building, 
has been postponed and will be rescheduled at a later date.

                                 

                * * * NOTICE--HEARING RESCHEDULED * * *

ADVISORY FROM THE COMMITTEE ON WAYS AND MEANS

                       SUBCOMMITTEE ON OVERSIGHT

                                                CONTACT: (202) 225-7601
FOR IMMEDIATE RELEASE
January 26, 2004
OV-10

                     Houghton Announces Hearing on

             IRS Efforts to Modernize its Computer Systems

    Congressman Amo Houghton (R-NY), Chairman, Subcommittee on 
Oversight of the Committee on Ways and Means, today announced that the 
Subcommittee will hold a hearing to evaluate the Internal Revenue 
Service (IRS) efforts to modernize its computer systems and hear the 
outcome of recent independent reviews of the IRS Business Systems 
Modernization (BSM) program requested by IRS Commissioner Mark Everson. 
The hearing will take place on Thursday, February 12, 2004, in the main 
Committee hearing room, 1100 Longworth House Office Building, beginning 
at 9:00 a.m.
      
    In view of the limited time available to hear witnesses, oral 
testimony at this hearing will be from invited witnesses only. 
Witnesses will include the Honorable Mark Everson, Commissioner of the 
IRS, a representative of the Computer Sciences Corporation (CSC), and 
Larry Levitan, Member, IRS Oversight Board.
      

BACKGROUND:

      
    United States taxpayers pay more than $2 trillion in taxes each 
year to the IRS. The IRS's receipt of tax revenue is dependent on a set 
of computer systems that has evolved over the past 35 years. Nearly all 
IRS employees depend on these computer systems to do their daily jobs, 
including more than 70,000 who use these systems to deliver direct 
service to taxpayers. The IRS network of computer systems is comprised 
of over 100,000 individual computers, 2,779 vendor supplied software 
products, and more than 50 million lines of IRS-maintained computer 
code.
      
    The existing IRS computer systems architecture became operational 
in 1967. Although significantly upgraded since then, the basic 
architecture and processing systems date from the 1960s. Since the 
1970s, the IRS has made numerous attempts at technological 
modernization. Success of past efforts was limited, due in part to the 
IRS's lack of a coordinated, unified plan and approach for fitting all 
aspects of modernization together.
      
    The current computer systems used by the IRS inhibit its ability to 
effectively carry out the mission of administering the Nation's tax 
laws. Maintaining these systems is a major undertaking and gets more 
difficult every year as older systems replacement parts and programmers 
are harder by which to come. Here are a few examples of problems caused 
by the outdated systems:
      
      American taxpayers expect a level of service from the IRS 
that rivals the private sector, but IRS systems do not provide the 
necessary functionality.
      IRS employees must work with taxpayer data that is not 
timely, resulting in frustration for both taxpayers and employees.
      IRS has trouble accounting for funds and maintaining the 
security of critical data and systems.
      
    Addressing these crucial needs, managing the inherent risks of 
modernization, and delivering the level of service taxpayers expect are 
all goals of the IRS's BSM.
      
    Prior efforts to modernize the IRS's computer systems in the mid-
1990s were not successful. These failures prompted the IRS to enter 
into an innovative contract with CSC in 1998 to launch the BSM program. 
Under the contract, CSC assists the IRS to design new systems and helps 
to identify contractors to perform software development and other 
tasks. The CSC is referred to as the prime contractor, under the 
principal BSM contract.
      
    Since mid-1999, more than $1.3 billion has been appropriated for 
BSM, including $391 million for FY 2002 and $366 million for FY 2003. 
The House and Senate have now approved an additional $390 million for 
FY 2004. The total cost of BSM is expected to be in the range of $8 
billion.
      
    Soon after his appointment, Commissioner Everson requested an 
independent review of a critical component of the BSM program to be 
conducted by Software Engineering Institute (SEI) of Carnegie Mellon 
University. The study focuses on the history of the Customer Account 
Data Engine (CADE) project and the feasibility of future plans with 
respect to CADE. The SEI has completed its interim review, and a 
witness from SEI will appear before the Subcommittee to present its 
findings and answer questions. The CSC has also commissioned Bain and 
Company to study CADE and other aspects of BSM and they will appear and 
be prepared to discuss Bain's findings. Finally, the IRS has conducted 
an internal review of the root causes of schedule delays and cost 
increases in BSM projects. The Commissioner is expected to discuss the 
results of all of the aforementioned studies and to announce changes 
that he plans to implement.
      
    In announcing the hearing, Chairman Houghton said, ``The IRS must 
bring its systems into the 21st century to provide the high and 
efficient level of service that our taxpayers expect. We saw this back 
at our 1998 hearing, and it's critical for us to see that this program 
is on track.''
      

FOCUS OF THE HEARING:

      
    The hearing will focus on the IRS efforts to modernize its computer 
systems and on independent reviews ordered by IRS Commissioner Mark 
Everson to assess the IRS's BSM program.
      

DETAILS FOR SUBMISSION OF WRITTEN COMMENTS:

      
    Please Note: Due to the change in House mail policy, any person or 
organization wishing to submit a written statement for the printed 
record of the hearing should send it electronically to 
[email protected], along with a fax copy to 
(202) 225-2610, by the close of business Thursday, February 26, 2004. 
Those filing written statements who wish to have their statements 
distributed to the press and interested public at the hearing should 
deliver their 200 copies to the Subcommittee on Oversight in room 1136 
Longworth House Office Building, in an open and searchable package 48 
hours before the hearing. The U.S. Capitol Police will refuse sealed-
packaged deliveries to all House Office Buildings.
      

FORMATTING REQUIREMENTS:

      
    Each statement presented for printing to the Committee by a 
witness, any written statement or exhibit submitted for the printed 
record or any written comments in response to a request for written 
comments must conform to the guidelines listed below. Any statement or 
exhibit not in compliance with these guidelines will not be printed, 
but will be maintained in the Committee files for review and use by the 
Committee.
      
    1. Due to the change in House mail policy, all statements and any 
accompanying exhibits for printing must be submitted electronically to 
[email protected], along with a fax copy to 
202/225-2610, in WordPerfect or MS Word format and MUST NOT exceed a 
total of 10 pages including attachments. Witnesses are advised that the 
Committee will rely on electronic submissions for printing the official 
hearing record.
      
    2. Copies of whole documents submitted as exhibit material will not 
be accepted for printing. Instead, exhibit material should be 
referenced and quoted or paraphrased. All exhibit material not meeting 
these specifications will be maintained in the Committee files for 
review and use by the Committee.
      
    3. Any statements must include a list of all clients, persons, or 
organizations on whose behalf the witness appears. A supplemental sheet 
must accompany each statement listing the name, company, address, 
telephone and fax numbers of each witness.
      
    Note: All Committee advisories and news releases are available on 
the World Wide Web at http://waysandmeans.house.gov.
      
    The Committee seeks to make its facilities accessible to persons 
with disabilities. If you are in need of special accommodations, please 
call 202-225-1721 or 202-226-3411 TTD/TTY in advance of the event (four 
business days notice is requested). Questions with regard to special 
accommodation needs in general (including availability of Committee 
materials in alternative formats) may be directed to the Committee as 
noted above.

                                 

    Chairman HOUGHTON. Good morning everybody. The hearing will 
come to order.
    The purpose of today's hearing is to examine the Internal 
Revenue Service's (IRS's) Business Systems Modernization (BSM) 
program. Five years ago, in late 1998, the IRS entered into an 
innovative contract with Computer Sciences Corporation (CSC) 
and a number of other companies to modernize the IRS's computer 
systems that had been designed back in the 1960s. This new 
program set out to avoid the pitfalls of an earlier failed 
effort, and it has produced results.
    Taxpayers for the first time last year were able to check 
the status of their refunds on the Internet, and small 
businesses could apply for an IRS identification number online.
    It is a huge program, and some of the key elements have 
experienced significant delays and cost overruns. These 
additional costs amount to $290 million, a sizeable percentage 
of the $1.7 billion appropriated for computer modernization 
through fiscal year 2004.
    Recognizing the problem, Commissioner Everson acted quickly 
to identify the source of these delays by commissioning several 
independent studies and digging into the details. Today, we are 
going to hear from the author of one of these studies, Steve 
Palmquist of Carnegie Mellon University, and from the U.S. 
General Accounting Office (GAO). In addition, we are going to 
hear from Larry Levitan, the Chairman of the IRS Oversight 
Board Committee that produced the review of the IRS computer 
system modernization. Finally, we are going to hear from the 
lead contractor, CSC, through the testimony of Paul Cofoni.
    Thank you for all being here, and I look forward to your 
testimony. Before I turn to my associate here, Mr. Pomeroy, I 
would just like to give a word of thanks to the extraordinary 
service of a former staff director of this Subcommittee, Mac 
McKinney. Mac is going to be leaving my office at the end of 
this month and has provided extraordinary service to us all 
during his 24 years on the Hill, 6 years with the Committee on 
Ways and Means, and 3 years as my chief of staff. Mac, where 
are you?
    [Laughter.]
    Congratulations and thank you so much. Now I would like to 
turn to Mr. Pomeroy.
    [The opening statement of Chairman Houghton follows:]

   Opening Statement of the Honorable Amo Houghton, Chairman, and a 
         Representative in Congress from the State of New York

    Good morning. The purpose of today's hearing is to thoroughly 
examine the Internal Revenue Service's Business Systems Modernization. 
Five years ago, in late 1998, the IRS entered into an innovative 
contract with Computer Sciences Corporation and a number of other 
companies to modernize the IRS's computer systems that were designed in 
the 1960s.
    The current computer modernization program was designed to avoid 
the pitfalls of an earlier failed effort, and has produced results. 
Taxpayers, for the first time last year, were able to check the status 
of their refunds on the internet, and small businesses could apply for 
an IRS identification number online.
    But, it is troubling that some of the key elements of the 
modernization program have experienced significant delays and cost 
over-runs. To date, these additional costs amount to $290 million, a 
sizable percentage of the $1.7 billion appropriated for computer 
modernization through Fiscal Year 2004.
    Recognizing the problem, Commissioner Everson has acted quickly to 
identify the source of these delays by commissioning several 
independent studies. Today we are privileged to hear from the author of 
one of these studies, Steve Palmquist of Carnegie-Mellon University and 
from the General Accounting Office. In addition, we will hear from 
Larry Levitan, the chairman of the IRS Oversight Board committee that 
produced a very thoughtful review of the IRS computer modernization. 
Finally, we will hear from the lead contractor, Computer Sciences 
Corporation, through the testimony of Paul Cafoni. Thank you all for 
being here, and I look forward to your testimony.
    I am now pleased to yield to our ranking Democrat, Mr. Pomeroy.

                                 

    Mr. POMEROY. Thank you, Mr. Chairman. First, let me echo 
your praise of Mac McKinney, who I have just enjoyed knowing, 
working with, and who represents, in my opinion, the finest 
dimensions of what the professional staff that keep this place 
running really represent. Mac, best wishes to you.
    Commissioner, I appreciate very much your leadership. I 
think that, speaking as Ranking Member of the Subcommittee on 
Oversight, the efforts you have made to reach out and keep me 
fully informed with your management initiatives, I have really 
appreciated, and I have a lot of respect for what you are 
achieving. We have seen, in fact, the role of computer 
technology giving a level of service to taxpayers beyond what 
they have seen before. I am particularly pleased they can now 
track the status of their refunds on the computer.
    At the same time, I think we can all acknowledge that the 
computerization of the IRS has certainly not necessarily kept 
up with computerization of other vast enterprises, public or 
private, but I would especially say we have been eclipsed by 
what has occurred in the private sector. I think that we all 
need to learn by what has happened. We have invested a lot of 
money, we have had consultants everywhere, and yet the progress 
may be not as far along as we would have hoped.
    So, I will look forward in the course of this hearing to 
learn from you, Commissioner, as well as some of the 
contractors about their thoughts in terms of how this is coming 
and how we might do it better. I think this is square in the 
strike zone of where the Subcommittee on Oversight has to pay 
its attention. We need to understand what the Commissioner's 
major initiatives are, how we can improve service of the IRS to 
the taxpayers of this Nation, and we need to be prepared to 
help the Commissioner, as well, with such focus and maybe 
additional funding as may be required. Thank you, Mr. Chairman, 
for this hearing.
    [The opening statement of Mr. Pomeroy follows:]

 Opening Statement of the Honorable Earl Pomeroy, a Representative in 
                Congress from the State of North Dakota

    The 2004 tax return filing season is well underway. Nationwide, 
taxpayers are in the process of filling out and filing their federal 
income tax returns with the Internal Revenue Service (IRS). IRS's 
computer systems have improved dramatically in recent years. Millions 
of taxpayers annually now file their tax returns electronically through 
E-File or Free File resulting in more error-free returns and quicker 
turnarounds for tax refunds. A taxpayer can even track the status of 
his or her tax refund check on the Internet to learn exactly when it 
will be mailed or deposited directly into a bank account.
    Taxpayer services provided to millions of Americans each year have 
been greatly improved through upgrades to many of IRS's automated 
systems. IRS employees now have the tools and information needed to 
resolve a taxpayer's problem while the person is on the telephone with 
the IRS. Similarly, the IRS is able to direct tax inquiries quickly to 
employees with special expertise which significantly reduces 
unnecessary waiting times.
    I remain concerned, however, that the IRS has a long way to go 
toward modernizing its overall computer systems architecture. With a 
system of 100,000 individual computers, nearly 3,000 vendor-supplied 
software products, and more than 50 million lines of computer code, it 
is obvious that the IRS must have a focused, coordinated, and unified 
plan for the 21st Century. I will be interested in exploring the 
agency's progress in this regard during today's hearing.
    I am particularly interested in learning more about the IRS's 
apparent inability to deliver on the ``big computer projects'' 
involving the master file of IRS records and system-wide infrastructure 
systems. I have learned that problems include inadequate performance by 
contractors, overly ambitious project portfolios designed by IRS 
management, and insufficient direct participation by IRS employees in 
the management of technology programs.
    I look forward to the Subcommittee's discussion of these issues and 
thank Chairman Houghton for scheduling today's important hearing.

                                 

    Chairman HOUGHTON. Thanks. Again, thank you, Commissioner, 
for being here. We are honored by what you are doing and 
helping all of us think through this maze of computerization. 
We look forward to your testimony.

   STATEMENT OF THE HONORABLE MARK W. EVERSON, COMMISSIONER, 
                    INTERNAL REVENUE SERVICE

    Mr. EVERSON. Thank you, Mr. Chairman, Mr. Pomeroy. I 
appreciate the opportunity to testify this morning on the 
status of the BSM program at the IRS. I also want to thank the 
Subcommittee for your continuing support of the IRS.
    The BSM is a key part of our broader agenda at the IRS. 
Before offering more details about the modernization program, 
let me set the stage with a few comments on our priorities at 
the IRS and the challenges we face.
    As you know, I have set three priorities for the IRS during 
my 5-year term as Commissioner. First, we must continue to 
improve service, making it easier for the taxpayer to 
understand and comply with the tax laws. Through focused 
implementation of the IRS Restructuring and Reform Act (RRA) of 
1998 (P.L. 105-206), the IRS has measurably improved service to 
taxpayers and practitioners. We aren't backing away from this 
commitment to service.
    The second area of emphasis is the subject for today, 
information technology. I will talk about that more in a 
moment.
    The third focus is to strengthen the integrity of the 
Nation's tax system through enhanced enforcement activities. As 
you know, the President recently transmitted the 2005 budget 
request to Congress. It calls for a 5-percent overall increase 
for IRS, including a 10-percent boost to our enforcement 
activities.
    We have four enforcement priorities. They are to discourage 
and deter noncompliance, with emphasis on corrosive activity by 
corporations, high-income individuals, and other contributors 
to the tax gap; assure that attorneys, accountants, and other 
tax practitioners adhere to professional standards and follow 
the law; detect and deter domestic and offshore-based tax and 
financial and criminal activity; and discourage and deter 
noncompliance within tax-exempt and government entities and 
misuse of such entities by third parties for tax avoidance and 
other unintended purposes.
    The budget request addresses each of these priorities, with 
which you are well familiar from your oversight work, and I am 
hopeful that you will be able to actively support our proposal 
for this increased funding.
    Now, let me turn to information technology modernization. 
As you have indicated, the IRS has made progress on major 
technology applications that provide enhanced services to 
taxpayers and practitioners. Examples include improved 
telephone service, electronic filing, and a suite of e-services 
to tax practitioners, but we have failed thus far to deliver 
big projects on the master files and infrastructure systems, 
the focus of our discussion today.
    Four studies completed last year consistently identified 
the following problems in delivering these large efforts: 
insufficient participation in the technology program by IRS 
business units, an overly ambitious portfolio, and inadequate 
performance by the contractor. The IRS is responding by 
increasing business unit ownership of these projects, resizing 
the project portfolio, reducing the modernization program from 
$388 million in fiscal year 2004 to $285 million in the 
President's 2005 request before the Congress now, and revising 
our relationships with the contractor and assuring joint 
accountability.
    It is this last subject that I want to speak to now. I 
would like to read to the Subcommittee a letter which I sent 
yesterday to Mike Laphen, who is the President and Chief 
Operating Officer of CSC.
    ``Dear Mike, I want to express my appreciation for your 
personal and active participation in improving CSC's support of 
the IRS's modernization effort. I believe that the IRS and CSC 
have made progress since we began our monthly meetings last 
August in identifying and addressing the challenges that 
confront us in modernization. The progress we have made makes 
it all the more disappointing that CSC has indicated that it 
will be unable to meet its revised delivery date of April 2004 
for the Integrated Financial System (IFS) project.''
    ``While I appreciate CSC's candor and your agreement to 
conduct the remaining work on the first release of IFS under a 
capped price arrangement at no additional cost to the IRS, I 
believe that our joint accountability for advancing 
modernization requires me to take action in response to this 
unsettling development. Accordingly, I have decided to direct 
our upcoming enforcement modernization projects for collection 
contract support and filing and payment compliance to other 
contracts.''
    ``As you can understand, I am not taking this step lightly. 
While no doubt unwelcome to CSC, I hope that this decision will 
lead to a sharpened focus and discipline and will, in fact, 
enhance the prospects for successful and timely delivery of 
other modernization projects by CSC. I see this approach as 
similar to what the IRS has done, as reflected in the 
President's 2005 budget request, in limiting the modernization 
portfolio in order to allow us to sharpen our focus and move 
more expeditiously.''
    ``In addition to the decision indicated above, we will 
carefully assess CSC's performance on current projects and the 
results of CSC's overall program management and integration 
efforts before awarding any follow-on work for existing 
projects. I look for CSC to demonstrate success on existing 
projects as it has recently on the important e-services suite 
of applications (which is significantly enhancing direct 
services to taxpayers and practitioners), in order to 
participate in new IRS modernization projects. In short, we 
need consistent, high-level performance and service from CSC in 
order for both sides to benefit from our partnership.''
    ``Again, thank you for your personal attention to these 
issues and CSC's acceptance of responsibility for this most 
recent delay. I look forward to working together to implement 
our modernization program successfully. Sincerely, Mark W. 
Everson.''
    This letter accurately summarizes where we stand. I want to 
assure the Subcommittee that the challenges facing the 
modernization effort are receiving my full attention. Thank 
you, and I would be pleased to take any questions and say good 
morning to Mr. Portman.
    [The prepared statement of Mr. Everson follows:]

  Statement of the Honorable Mark W. Everson, Commissioner, Internal 
                            Revenue Service

Introduction And Summary
    Chairman Houghton, Ranking Member Pomeroy, and distinguished 
Members of the subcommittee, I appreciate the opportunity to testify 
this afternoon on the status of the Business Systems Modernization 
(BSM) program of the Internal Revenue Service--and to discuss our 
recent review of the program's health. I also want to thank you for 
your support in working with the House Appropriations Committee to 
ensure that the IRS receives adequate funding to carry out the program.
    Business Systems Modernization is a key part of our broader agenda 
at the IRS. Before offering more details about the modernization 
program, let me set the stage with a few comments on our priorities at 
the IRS and the challenges we face.
    At the IRS, our working equation is service plus enforcement equals 
compliance. The better we serve the taxpayer, and the better we enforce 
the law, the more likely the taxpayer will pay the taxes he or she 
owes.
    To support this philosophy of service plus enforcement equals 
compliance, we are guided by three themes.
    First we are improving service, making it easier for the taxpayer 
to understand and comply with the tax laws. We have divided the IRS 
into ``customer segments''--including wages and income, small, medium 
and large businesses, non-profits. In the last four years, our toll-
free telephone service has risen sharply. Downloads of IRS forms from 
our website has soared. Electronic filing of taxes has jumped from 29 
million in 1999 to 52 million last year and nearly half of all 
taxpayers are expected to efile this coming year.
    Second, we are boosting enforcement, a key emphasis of the 
President's 2005 IRS budget request just sent to Congress.
    By our best estimates, we lose a quarter trillion dollars each year 
because taxpayers do not pay their tax voluntarily or in a timely 
fashion. (This is a rough estimate based largely upon 1988 data from 
our old Taxpayer Compliance Measurement Program).
    Over the last four years, the number of Americans saying it is OK 
to cheat on taxes rose from 11 to 17 percent. Sixty percent of 
Americans believe that people are more likely to cheat on taxes and 
take a chance on being audited.
    This drop in compliance coincides with drop in enforcement of the 
tax law. Since 1996, the number of IRS revenue agents, officers, and 
criminal investigators has dropped by over 25 percent.
    At the IRS we have begun to address the tax gap crisis. We have 
shifted badly needed resources so we can hire more front-line 
enforcement personnel--who will primarily focus on non-compliance among 
high income individuals and businesses.
    In addition, I am most pleased and grateful that the President's FY 
2005 budget submission requests an additional $300 million for 
enforcement activities over the FY 2004 consolidated appropriations 
level.
    What will this extra $300 million do?
    It will help to:

      Discourage cheating and non-compliance, particularly by 
corporations, high income individuals and tax exempt groups.
      It will help attorneys, accountants and other tax 
professionals adhere to professional standards and obey the law.
      It will detect and deter domestic and off-shore tax and 
financial criminal activity.
      Discourage and deter non-compliance within tax-exempt and 
government entities and misuse of such entities by third parties for 
tax avoidance and other purposes.

    Our third focus is modernization of our information technology, 
often referred to as Business Systems Modernization or BSM. Most of our 
tax administration systems are very old and difficult to keep current 
with today's fast paced environment; they must be modernized.
    We are committed to resizing our modernization efforts to allow 
greater management capacity and to focus on the most critical projects 
and initiatives. We used comprehensive studies over the summer to help 
us identify opportunities to improve management, re-engineer business 
processes and implement some new systems and technology. The FY2005 
budget provides $285 million to continue this effort to replace current 
business systems and technology.
    However, equally important is the critical role technology 
modernization plays in enabling customer service and enforcement goals. 
To the IRS, ``modernization'' is broader than just those parts funded 
by BSM. To the taxpayer it is the full suite of technology enabled 
services that we provide, such as irs.gov website and telephone 
technology.
    ``Modernization'' also includes projects with which taxpayers would 
not be directly involved, such as replacing our master file system, 
implementing the on-line security features, and building the modernized 
technological infrastructure on which all of our future modernization 
applications will build and depend.
    To date, after five years, we have achieved mixed results with the 
modernization program.
Program Challenges
    One aspect of BSM on which there was unanimity is that this program 
is as complex and challenging as any information technology program in 
the world. When nominated last February, I began learning about the 
current Modernization program from both IRS executives and others 
inside and outside the government. I learned of earlier unsuccessful 
attempts to modernize the IRS. But I also learned that staff was 
optimistic they were beginning to control the difficulties encountered 
in the first four years. We expected to achieve some key milestones in 
the summer and fall, such as the initial delivery of the system that 
would start to replace our antiquated tax accounting system, called the 
Customer Account Data Engine (CADE).
    Before the summer was through, it became clear two very significant 
modernization projects, CADE and the Integrated Financial System (IFS), 
would experience substantial delays. These were two major setbacks.
    These delays were clear evidence that significant problems still 
existed in the BSM program. Based on both external and internal 
assessments, it became apparent to me that we needed to address four 
key elements needed in order to turn the program around:
    First, the scope of the projects was far too large. After working 
very closely with the Treasury Department and the Office of Management 
and Budget (OMB), and evaluating recommendations we received from the 
Oversight Board and Congress, we narrowed the scope and number of 
modernization projects. The IRS management team and the PRIME 
contractors had taken on too much and been stretched too thin. We did 
not have the capacity to properly manage such a large portfolio. The 
result is that we have been unable to devote the resources, energy and 
attention to meeting our primary goals.
    I am pleased to report that over the last year, we made progress in 
this area. We considerably reduced the size and scope of the 
modernization program, and have been working with the Department and 
OMB to seek more opportunities to better balance management capacity 
with the modernization portfolio, without dramatically reducing the 
program's effectiveness.
    Second, a much greater degree of business ownership and 
participation was critical. The modernization program will only be 
successful if the most senior and experienced IRS business leaders take 
ownership of the program.
    While we have a great deal of work to do, we have made some 
progress in this area as well. I appointed John Dalrymple, who has 
spent over 30 years focused on front-line taxpayer issues, as the 
Deputy Commissioner for Operations Support. In addition, I appointed 
the former IRS CFO, Todd Grams, to the CIO position to bring better 
management and financial discipline to the technology modernization 
program.
    These management appointments represent a change in the way we have 
pursued modernization projects in the past. These appointments are only 
the beginning of bringing more management discipline and increased 
business involvement to our modernization efforts. Success in this area 
will require a willingness on our part to deviate from past practices, 
including a change in the allocation of accountability for 
modernization projects. We will appoint the appropriate people to these 
critical projects, clearly set expectations, and hold these people 
accountable for the results.
    Third, we needed significant improvements in the performance of our 
PRIME contractor. It's no secret that our projects have consistently 
run late, delivered less functionality than planned, and cost 
significantly more than targeted. While the PRIME has improved its 
performance, nevertheless, delays and cost increases persist, as 
evidenced by the recent slippages in CADE and IFS deliveries. We need 
to have a PRIME contractor that consistently meets its commitments.
    Fourth, we needed fresh and independent assessments from outside 
experts on the health of the modernization program, as whole, as well 
as specific projects, such as CADE.
    All of the assessments confirmed that the IRS modernization effort 
is a massive, highly complex, high-risk program that is confronting a 
number of critical management and technological challenges. These 
studies also made it clear that we should not turn back, but rather 
make a series of changes to strengthen our current program. While all 
of these studies assessed different components of the program, they 
clearly suggested consistent improvement opportunities.
    Looking beyond the conclusions drawn for the overall program, the 
outside assessments concluded we need significant improvements in 
execution to put our most critical projects back on track.
    I do not intend to understate to the Committee the seriousness of 
the current challenges faced by the modernization program or my 
awareness of the long history and inconsistent record of success on 
modernization. But I do think it is worth noting that we have achieved 
some significant accomplishments to date.
Delivering Benefits
    First, we have developed a broad strategic plan called the 
Enterprise Architecture, encompassing both the functional and 
technological dimensions of the BSM program. We recently won the 
``Excellence in Enterprise Architecture Leadership'' award for this 
Enterprise Architecture from E-Government.
    Second, we established a secure, on-line technical infrastructure 
to support both new BSM applications as well as other future IRS 
applications.
    Third, we have delivered applications that provide tangible 
benefits to taxpayers and improve the efficiency and effectiveness of 
our tax administration system. They include:

      Customer Communications, which reduced taxpayer call-
waiting time in half, reduced the number of abandoned calls by 50%, and 
doubled the number of refund inquiries from our Spanish speaking 
taxpayers.
      Where's My Refund?/Where's My Advance Child Tax Credit?, 
which gives taxpayers instant updates on the status of their tax 
refunds and advance child tax credits. These applications have received 
over 40 million requests since the beginning of the year. By shifting a 
significant volume of customer demand to the Internet, we have seen a 
measurable improvement in service to taxpayers who still choose to 
call.
      e-Services, which includes preparer tax identification 
number (TIN) applications with instant delivery, individual TIN 
matching for 3rd party payers, on-line registration for electronic e-
Services, and on-line initiation of the electronic originator 
application (currently released to a controlled segment of external 
users). I am pleased to announce that we recently made the first part 
of e-Services available on our public web site. The remaining parts 
will come out over the next several months.
      Internet EIN, which allows small businesses to apply for, 
and receive, an Employer Identification Number on-line. Since its 
launch last May, we have processed over 453,000 EIN applications that 
have come in over the Internet.
      HR Connect, which allows IRS users to perform many 
personnel actions on-line. To date, we have rolled this out to about 
73,000 employees, and will complete the rollout by early 2004. This 
technological advancement will enable the Service to redirect hundreds 
of positions to enforcement activities by the time it is fully 
deployed, which we have planned for October 2005.

    Now, let me provide a summary of the programmatic reviews, and the 
current program status, beginning with CADE.
Carnegie Mellon Software Engineering Institute (SEI) Assessment of CADE
    CADE, as most of you already know, is designed to replace the IRS' 
current Master Files. The Master Files are the Service's central and 
official repository of taxpayer information. As such, it is a 
singularly important system to the IRS. The Master Files are 
drastically over age in the life cycle of technology, having been in 
place since the 1960s. They are considered reliable but are extremely 
inflexible and fragile, and therefore difficult to support, in part 
because of the old technology they use.
    We initially planned CADE to be available for the 2002 filing 
season, but we missed several dates. Following the acknowledgment in 
August 2003 that we would not meet the latest scheduled release of the 
first component of the CADE project, I requested that the Carnegie 
Mellon Software Engineering Institute, or SEI, provide an independent 
assessment of CADE's history and the feasibility of future plans for 
CADE.
    SEI advised us to stay the course with this first release of CADE, 
which would handle 1040EZ returns for single filers that were either 
even balance or which had a refund due. They felt the architecture for 
this release was sound.
    However, they strongly urged us to start paying more attention to 
two areas. One was to get started on building a version that would 
utilize the future technology for CADE, which the first release does 
not. The other was to accelerate plans to better understand the scope 
of future releases to ensure that our approach was sound for the long 
term.
    We have taken steps to address both of these recommendations and we 
plan to have SEI come back periodically to check on our progress.
    Let me briefly update you on where we are with CADE.
Status Of The CADE Project
    First, we have now passed some critical hurdles with the version of 
CADE that would have gone into production last summer, such as the 
system integration test, a pilot, and a system acceptance test. PRIME 
will address the defects identified during this process in the release 
currently under development.
    Second, we are well along on implementing the changes needed to 
create the 2004 filing season version of CADE so that we can process 
actual tax year 2003 returns in 2004. PRIME has developed a detailed 
plan to be able to develop this ``working CADE'' by this summer. While 
that will clearly be too late to process any significant number of tax 
year 2003 1040EZ returns, it will enable us to launch CADE in a low 
risk setting and gain valuable operational experience.
    Third, we are taking steps to ensure that the future changes needed 
to deploy CADE for the full 2005 filing season will be tested and ready 
by January 1, 2005.
    Let me also note that the vast majority of the work on the first 
component of CADE is being conducted under a type of contract which 
limits the IRS's PRIME cost to their original proposed costs for the 
first release. Additional costs are, however, being incurred for filing 
seasons changes that we now need to apply due to the delay.
Other Modernization Assessments
    In addition to the assessment of CADE undertaken by SEI, we 
commissioned two other outside, independent assessments of various 
aspects of Modernization. These studies include the PRIME Review 
assessment, conducted by Bain and Company, and Assessment of the IRS 
Office of Procurement, conducted by Acquisition Solutions, Inc. The 
purpose of the PRIME Review was to identify causes of breakdowns in 
business processes and the engagement model and to provide recommended 
solutions. This assessment included identification of specific skill 
and leadership gaps, an audit of CADE, and finalization of the 
requirements definition and validation of the Integrated Master 
Schedule.
    While the primary purpose of the Assessment of the IRS Office of 
Procurement was to determine the efficiency and effectiveness of the 
IRS' Procurement organization, it did provide a brief assessment of the 
Modernization contracting program.
IRS Responses to Internal and External Assessments
    First let me say that we have accepted the recommendations made by 
these external reviewers and are moving aggressively to implement them.
    I have significantly enhanced the top-level leadership of the 
modernization effort. I personally meet with the IRS top team every 
other week to discuss the status of critical project goals and other 
program improvement initiatives.
    I have also expanded my personal interactions with the PRIME 
Contractor, Computer Sciences Corporation (CSC), by increasing my 
personal engagement with their top-level executives. I meet monthly 
with the President/COO of CSC to provide clear direction on performance 
expectations and accountability.
    In addition to strengthening the top team, I have asked John Duder, 
the Deputy Commissioner for the Wage and Investment Division, to 
transfer to the modernization program and focus his complete attention 
to delivering the CADE project. I have also asked Rich Morgante, the 
Deputy Commissioner for the Tax Exempt and Government Entities 
Division, to join the modernization program to lead the implementation 
of the recommendations made by the external reviews.
    Now that I have discussed the Modernization program's 
accomplishments and the improvements underway, let me address the 
status of our most critical projects.
Status of Other Tax Administration Modernization Projects
    I have discussed CADE at some length because, like the new online 
infrastructure that we have deployed, it is a key foundational 
component of the modernized systems. I will only briefly discuss the 
other four systems that we are developing, two of which address tax 
administration and two of which address internal financial management.
    The two new tax administration systems are e-Services and 
Modernized e-file.
    e-Services provides an array of services aimed at larger third 
party providers, including electronic return originators and 
institutions that report 1099's, such as banks and brokerage firms.
    The e-Services suite of projects is over a year late and its cost 
has increased almost 100% over estimates made several years ago. Given 
that these are the first applications to use major components of the 
new modernized infrastructure, and given the complexity of linking to 
our very old current systems, it is not surprising that we did not meet 
our initial estimates for costs and schedule; however, we must adhere 
more closely to cost and schedule estimates in the future.
    I am very pleased to inform the Subcommittee, therefore, that we 
now see sunshine at the end of the e-Services tunnel. We have deployed 
the first parts of a multipart release, including online registration, 
interactive Taxpayer ID Number Matching, and online Preparer Tax ID 
Number application, and made them available on irs.gov for several 
months. In addition, on January 28, 2004, we fully deployed the 
electronic return originator application on irs.gov. Over the next 
couple of months, we will deploy additional functions, like electronic 
account resolution, transcript delivery, and online Disclosure 
Authorization applications.
    Modernized e-file will provide electronic filing for the first time 
to large corporations and tax exempt organizations.
    The Modernized e-file application achieved a very significant 
milestone last November when we began software certification testing on 
schedule for dial-up users. This certification is the process by which 
we validate that vendor software products work properly with our 
systems. On February 4, 2004, we began certification testing over the 
Internet.
    We had hoped to open Modernized e-file for business at the 
beginning of January, but our target date has slipped to mid-February. 
Later releases of Modernized e-file will enable the filing of 
additional 1120 and 990 schedules. Eventually, this project will 
replace our almost 20-year old 1040 e-file system.
Status of Internal Management Modernization Projects
    The two internal projects we are developing are the new Integrated 
Financial System (IFS) and the Custodial Accounting Project (CAP), both 
of which are disappointments at this time.
    We are working on internal systems to correct and address a number 
of financial material weaknesses that reflect internal management 
deficiencies, such as compliance with current federal accounting 
standards and the ability to accurately report on tax-related income. 
While we are very proud to be able to close our books on the completed 
fiscal year by mid-November, and to have received a clean audit opinion 
from GAO the last three years, nevertheless, we are operating with 
deficient internal management systems that urgently need correction. We 
have tried to balance the work on tax administration systems with the 
necessary work on internal management projects within the Modernization 
program, though, as I noted earlier, there is no doubt that we have 
been trying to do too much overall.
    In the case of IFS, we believe that we understand the causes that 
led us to miss the target production date of October. We initially set 
a target for IFS that was, in the end, too aggressive, and we did a 
very poor job of setting expectations and communicating our risk. On 
January 30th, the CSC executives informed us that their revised target 
of April 2004 was not achievable. The new target date is October.
    This further delay in IFS delivery is a huge disappointment both to 
the IRS and to me personally. While the PRIME has indicated their 
willingness to bear the financial burden for this further delay, we 
feel that the IRS needs to take some stronger steps. Specifically, we 
will expand the competition for the new enforcement projects that we 
plan to start later this year and next year. In addition; we will also 
expand the competition for the next phase of IFS.
    CAP is also a complex project. Whereas, for IFS, we are using a 
standard Commercial-off-the-shelf (COTS) product that we are tailoring 
to the IRS, CAP is largely custom-developed software tied to the 40-
year old Master Files system. Over the past several months, progress 
has slowed significantly as the CAP project drives through hundreds of 
millions of records in the Individual Master File.
    After intensive analysis of the factors causing the slowdown of 
progress on CAP, we have taken a number of steps that I believe will 
get CAP on a more predictable track. We now expect to begin loading the 
CAP data warehouse in May to ready it for production in August. In 
addition, we are currently negotiating an agreement to limit the 
government's financial risk with the CAP contractor, Northrop Grumman 
Mission Systems, for the first release of CAP.
    In order to get a jump on the second release of CAP, we started 
some design work last April, but we have now decided to significantly 
delay that project, both to ensure total focus on the first release and 
to wait until we are certain that the approach to CAP in the first 
release is what we want to use in the second release.
Conclusion
    While we have much work to do on the modernization to meet all of 
the challenges we currently face, I can assure you that it is one of my 
top priorities as Commissioner. I also assure you that we have a solid 
oversight relationship with Treasury, OMB, Congress, and the IRS 
Oversight Board. We are working closely with these stakeholders to 
ensure they are well informed of program goals and the status of the 
projects against schedule and cost targets. All parties involved in 
modernization are keenly aware that our first goal is to ensure we 
spend taxpayers' dollars wisely.
    The reviews that the PRIME and I commissioned, and the actions we 
are now taking, reflect our strong commitment to get Modernization 
``right.'' The long-term future of the tax system depends so much upon 
our success. While the progress to date has been decidedly mixed, we 
need to put in place the foundation upon which the tax system will 
build and rely for decades to come.
    Mr. Chairman, Ranking Member Pomeroy, and distinguished Members of 
the subcommittee, again let me say that I appreciate your leadership 
and the continued support of the subcommittee, and would be pleased to 
answer any questions you may have.

                                 

    Chairman HOUGHTON. Would you like to make an opening 
statement?
    Mr. PORTMAN. No thanks.
    Chairman HOUGHTON. Okay. Obviously, this is a critical 
issue, critical for everybody, not just the mechanics of the 
IRS, and I think it makes a lot of sense having a little 
competition. Of course, the question is how long will it take 
for enforcement and other things to get up to speed because of 
the total package involved.
    Let me ask you a question. When you have these contracts, 
Commissioner, do you have anybody from the IRS working inside 
the contractor's shop as well as having them working inside 
your shop to get a feel of timing, meeting deadlines, because 
when you take a look at the span of years such as we have had 
with this $1.7 billion contract, with its terrific overrun, 
about 17 percent overrun, how did they get so off track?
    Mr. EVERSON. There are several points you have in there, 
Mr. Chairman, and let me take them in order. I agree with you 
about the point you make about the action we are taking, the 
steps we took yesterday. I have run businesses and had accounts 
where you are servicing 100 percent of the supply to the 
customer. You work as hard as you can to protect that position, 
but when the day comes that it is opened up and there is a 
second supplier, it changes everything. In my experience, it 
usually makes you a better supplier because you are held to two 
standards, not just your own standard of your relationship with 
the customer, but you also are being compared constantly 
against your competition.
    This is not a small step. In the $285 million request that 
we have pending before the Congress, only $40 million of it 
pertains to the enforcement modules that I mentioned. The CSC 
will still participate in a portion of that because they have 
the overall integration role, but what we are suggesting is we 
will look at other contracting vehicles. So, your first point 
is entirely correct. This is a significant step. What we tried 
to craft was a balanced approach.
    I talked yesterday with Mr. Laphen, who has been terrific 
to work with through all of this. He understands. He 
characterized the action that we took as balanced and providing 
incentives as CSC is able to meet commitments and demonstrate 
improvements to gain the future work, without taking an overall 
too harsh step at this time, which would be in nobody's 
interest given how far we have come. So, that is the first 
point I would suggest.
    You asked about how this program gets off kilter over a 
period of time and we get to these overruns. I think that I 
would go back to those three conclusions that the four somewhat 
overlapping studies indicated, and you are going to hear from 
one of the groups that helped us on this.
    Again, I don't think that the IRS had adequate business 
unit oversight or ownership of these projects. They were 
treated as technical projects driven by the technical staffs 
and by the contractor. That means that our business unit people 
didn't help set the standards. They weren't in there at an 
early enough stage working on the testing, making sure that 
things were happening.
    As a result, we got a very cumbersome system where just a 
simple change request took 30 or 40 weeks to get through. This 
adds incredible cost because you are circling back and redoing 
work, and you haven't anticipated up front what you need to do.
    So, we are looking at all that. We are tightening it up by 
having these projects driven out of the business units. As one 
example, you are all familiar with our structure that came in 
through RRA 1998. We have got this Wage and Investment group 
that is for the bread and butter taxpayer, the 100 million-plus 
returns that come in each year. Our Deputy Commissioner for 
that whole operation is now in charge of the Customer Account 
Data Engine (CADE) project because they are going to be the 
users. So, CSC and our technical staff work for John Duder, who 
is the deputy there. That is a big change. That will help us.
    Lastly, on the very specific point of co-location, it is 
something that Mike Laphen and I have talked about. As we 
started to work together, one of the observations we had was 
just what you said, that we have got too much independence 
here. We have talked about having a lot more co-location of our 
staffs and CSC is doing that, and we are working with them. 
Mike and I haven't agreed to have a co-located office, but it 
may come to that, you never know.
    Chairman HOUGHTON. Thanks very much. I know with your 
ability and your background experience, this will work out 
well. It really, at some point it has to work----
    Mr. EVERSON. Yes.
    Chairman HOUGHTON. We have had so many approaches here that 
have fallen short, every time, and the whole system is banking 
on your success.
    Mr. EVERSON. It is, but I do want to reassure you. As you 
have indicated, I give the IRS a mixed grade here. A lot has 
been done. I don't want people to tar the whole effort here. As 
I meet people around the country, they say, boy, some of these 
things you are doing with the technology are just great. It is 
largely these internal systems that we just haven't been able 
to crack the nut on yet.
    Chairman HOUGHTON. Thank you very much. Mr. Pomeroy?
    Mr. POMEROY. Thank you, Mr. Chairman. Commissioner, I found 
your testimony very interesting and typical of what our 
relationship has been, very forthright on your part in 
identifying problems both with the IRS and with the contractor. 
I think that it is a lesson that is learned every day in terms 
of the relationship between consultants and the enterprise, 
public sector, private sector, wherever. You have to have very 
focused relationships and you have to have an awful lot of 
involvement or the thing is going to get away from you. It 
sounds like it got away from us a bit on especially the main 
computer project, is that your assessment?
    Mr. EVERSON. I think everybody had the best of intentions 
here, absolutely a real desire to be successful. The very scope 
and complexity of these projects doesn't give you much of a 
choice. You either get into it and understand it, or it is 
almost too much for you.
    Mr. POMEROY. Yes.
    Mr. EVERSON. What I think happened was the people who 
needed to understand these projects, meaning the folks that run 
our operations, were able to take a pass because of the way we 
structured it.
    It would be easy for me to sit here and just trash the 
contractor. I am not doing that. There was a joint failure to 
deliver in which the IRS clearly shares. I don't think we were 
correctly configured and managing our efforts. So, it did get 
away from us, as you indicated.
    Mr. POMEROY. I think that as government looks at a greater 
role of outsourcing right across the board, this is something 
we need to keep very much in mind, and you have to have on 
staff requisite expertise to really know these projects, to be 
a part of the projects, to oversee the projects. If we are 
going to slash staff internally and outsource and somehow give 
the taxpayer a better value, that isn't always going to work, 
especially if you are unable to sufficiently be involved in 
monitoring the project.
    Mr. EVERSON. I couldn't agree with you more. When you look 
at business processes, whether they are your own people or they 
are automated or they are external people or goods or services, 
you have to treat them all as if they were in your own shop and 
work very closely together. That is a challenge when you have 
outside services. That needs to be managed correctly.
    Mr. POMEROY. Is there mitigation that the Federal 
Government should initiate relative to nonperformance or 
violated contracted performance by the contractors?
    Mr. EVERSON. No, I don't think so. I think at this stage, 
our relationships with the contractor are good. I would suggest 
to you that the relationships, in fact, are very strong. It is 
just----
    Mr. POMEROY. We get our money back.
    Mr. EVERSON. If you can--I would love it if you could get 
some money back--ask Mr. Cofoni. Maybe he can give you some 
back. He will be talking later, but I think what CSC has done 
is indicated they are going to cap this IFS release, which is 
good news for us, making sure we won't incur additional 
overruns, as the Chairman indicated. We are going to be working 
very closely in looking at fixed-price alternatives as we go 
forward, so----
    Mr. POMEROY. I am pleased about that. It just would seem to 
me within your kind of fiduciary responsibility as manager, 
someone either within IRS or the U.S. Department of Justice 
ought to be evaluating whether or not there is a liability 
issue that ought to come back.
    Mr. EVERSON. We monitor the contracts closely and the 
preponderance of them, up to now, have been a cost-plus basis, 
so that if you mismanage the process, as we were just chatting, 
the costs go up.
    Mr. POMEROY. Okay. The final thing I would mention is I 
know that you and I have spoken about taxpayer compliance and 
how disappointed we have been that some very well established 
players in the marketplace have been repeddling tax shelters 
that really are not grounded very well under law and fail any 
test of ethics, of business ethics. What I am wondering is, as 
we try to look at adding structural disincentives for any 
enterprise that might consider this kind of conduct, is there 
cross-linkage relative to contractual relationships with the 
IRS? Now let me point my question.
    KPMG is a fine, long-established firm, one of the best in 
the world, but they had a renegade arm that somehow was very 
involved in peddling tax shelters to prospective clients that 
they had gone an awful long way, strayed a long way from the 
path of what they usually do in this activity. I am fairly 
stunned by it, and I think it is disgraceful. I think that they 
had a very bad day in the Senate hearing when this information 
was brought into public light.
    Is this the kind of thing where if they have a 
contractual--if one arm of that massive firm is doing that 
activity and another arm is doing legitimate contracting 
activity with the IRS, is that something you might evaluate 
relative to your continued participation with them as a 
contract entity in light of their corporate conduct at large?
    Mr. EVERSON. I think that is a question that is really 
probably best addressed to the Office of Management and Budget 
(OMB) and to the Office of Federal Procurement Policy because 
of their overall supervision of procurement practices. It is 
clearly legitimate for the government to hold Federal 
contractors to higher standards in some instances than would be 
the case between normal commercial relationships or parties in 
the private sector, and the government does do that. The 
government asks for additional information on certain areas--
have there have been Equal Employment Opportunity complaints, 
just as an example. There are considerations that enter into 
procurement decisions as they are taken.
    So, without getting into the real merits of what you are 
asking, which would obviously have to be fact-based 
considerations and couldn't rely on simply a hearing, no matter 
how compelling, you would have to have had some action that 
would have taken place administratively or in the courts before 
you would consider an action. The OMB does occasionally say, as 
it did with some of the big corporations that got into 
problems, like Arthur Anderson and others, that they are 
disbarred from participating in Federal contracts. So, that is 
looked at, but I would suggest to you it is more an OMB action.
    Mr. POMEROY. I think that answer--I know I am done, but I 
just have a summary point, Mr. Chairman. Your answer is a 
little too circumspect for me. In other words, I think that 
the----
    Mr. EVERSON. I thought you might react that way.
    [Laughter.]
    Mr. POMEROY. I think you have a role here. I am aware of a 
memorandum within this concern that assessed the potential 
profit from marketing these activities against the potential 
downside in terms of the structural fines, and they said even 
if we are fined, it is still a net plus for us. Well, we need 
to build in marketplace disincentives for those that don't 
respond on businesses' sheer honesty and honesty being the best 
policy. If they want to engage in this kind of conduct, we need 
to make certain that there is a business consequence that 
produces a net loss.
    Mr. EVERSON. Yes.
    Mr. POMEROY. So, beyond the fine, I would like these 
concerns to understand that their ability to do business with 
the Federal Government, the Federal Government that on the one 
hand they are ripping off, is forever placed--not forever, but 
for a period of time placed in great jeopardy.
    Mr. EVERSON. Mm-hmm.
    Mr. POMEROY. One aspect of their operation that is ripping 
off the government may jeopardize the full scope of business--
--
    Mr. EVERSON. Right.
    Mr. POMEROY. They may otherwise do with the government.
    Mr. EVERSON. Let me put a finer point on it. Our job, first 
and foremost, is to address the substance of the problems. If 
we feel that a professional services firm is violating promoter 
regulations or statutes, we have to address that, and we are 
addressing that in this area of abusive shelters. We have a 
very active set of programs, and where there has been 
misconduct, I believe that there will be consequences, and that 
is how I would leave it.
    Mr. POMEROY. Thank you. Thank you, Mr. Chairman.
    Chairman HOUGHTON. Mr. Commissioner, you know that this is 
a very bipartisan Committee. We walk in lockstep on many of 
these issues. We try not to be partisan. I am sitting between 
two of the finest Representatives we have in Congress, and I am 
really honored when I say this. I would like to then see if you 
can come up to this standard, to introduce Mr. Portman.
    [Laughter.]
    Mr. PORTMAN. I was looking to my right when he said that to 
see who he was talking about.
    [Laughter.]
    Thank you, Mr. Chairman, and thanks for your willingness to 
have this hearing and your continued oversight of the IRS. It 
is not something that Congress has done terribly well over the 
years, and I think with Mr. Houghton coming on the Subcommittee 
on Oversight, we have done a much better job of keeping track 
of what is going on and hopefully being a constructive partner. 
So, I thank you and your staff for continuing to do that.
    Here we are again. If you think about it, back in the mid-
1990s, the reason the Restructuring and Reform Commission was 
started was not so much about all the things we ended up 
recommending to do with management and policy as it was about 
modernization, because it really came out of the frustration 
felt by the Appropriations Committees over appropriating more 
and more money toward a modernization program at that time 
which was alleged to have spent $3 billion. It was to little or 
no benefit to the taxpayer or to the government.
    We are not at that level yet, but I look at the GAO 
recommendations and its analysis, and we are going to hear from 
them in a moment, and we are beginning to get to the point 
where we are talking about hundreds of millions. This is a 
constant frustration. I know it is very complicated. The GAO 
says there are $290 million of cost overruns, a cumulative of 
83.5 months of delay in major BSM projects.
    So, here we are again, and Commissioner, you are relatively 
new to this and therefore you have the ability both to look at 
it in a more objective manner, I believe, and also to be able 
to shake things up a little. It sounds like you are doing that. 
I think the recommendation of having the business units take 
direct leadership and ownership of the modernization projects 
is certainly a step in the right direction. My question is, is 
this more a management issue at the IRS or a contractor issue, 
and you have answered that earlier by saying you think it is 
both.
    With regard to the IRS problem, which is where this 
Committee has taken a big interest in the past in literally 
trying to restructure the IRS, and we have gone through that 
over the past few years and there has been a lot of disruption 
at the IRS in that process, has that disruption caused some of 
these problems? In other words, by changing our structure to 
focus in the three areas by literally creating new positions--
now you have two deputies, for instance--has that been part of 
the problem, that over this time period, the last 5 years, the 
IRS itself has been in somewhat of turmoil because of the 
changes. If that is part of the problem, what should we be 
doing about it?
    Mr. EVERSON. I think that the reforms that you took, 
creating business units, were sound. I think that remains the 
core of what you did in RRA 1998. I think that was a good 
reform.
    I do believe that what I did last June when I came in, 
which is to consolidate the staff functions under one deputy, 
was an important follow-on step to that because it was 
important that they cooperate and support each other. As I have 
said to many, in the private sector, the staff functions all 
work together because they don't have the same juice that the 
manufacturing guys or the marketing people do who deliver the 
bottom line. In government, they are all fractured and they are 
impotent, so that things like common sense controls or good 
technology or good human resource practices are lost against 
the program managers who are running the businesses.
    By pulling them together and bringing them under the 
leadership of our Senior Operations Manager, the man who was 
running the Wage and Investment Division, the perspective is 
that they are accountable to the operators now, which was not 
what the going model was.
    I don't think that the changes through the reorganization 
have contributed to this problem. In fact, I think that Charles 
Rossotti and the people in the IRS did a spectacular job of 
standing up the new organizations, and I admire everybody for 
the tremendous job they did, how relatively pain-free, from my 
perspective, that was accomplished, and the great improvements 
in service, with which you are familiar, that took place at the 
same time. A lot of times, you get a lag effect where service 
hasn't gotten better because you are consumed in the 
reorganization.
    The IRS did both. It reorganized and----
    Mr. PORTMAN. Improved service.
    Mr. EVERSON. It improved service. That is a phenomenal 
accomplishment. Now, as you know, we did it at the expense of 
enforcement. We are redirecting resources to that, and we are 
working on it, but now we are also bringing in closer these 
support functions. One of my obligations, I have talked to all 
three of you about it, is to make sure we run the organization 
efficiently and economically, and we are doing that as we 
restructure elements of the work force, and this technology is 
a piece of it.
    No, I don't think that there has been anything out of 
sequence here. I think it is a good sequence, and it is 
working.
    Mr. PORTMAN. To what do you attribute the management 
failure of this? Now, I am getting to the contractor issue in a 
moment, but in the Restructuring and Reform Commission, we 
actually found, as I recall, that the IRS was not contracting 
too much, and I understand Mr. Pomeroy's point of view on the 
outsourcing and the contracting concerns and the oversight 
concerns. Our concern, as I recall, was more based on analysis 
that the IRS was trying to do too much in-house. For example, 
there were some programs that were literally off-the-shelf 
programs that could have been used and instead the IRS insisted 
on recreating the wheel.
    We also found that the management structure, the culture of 
the management was risk averse. We found that there was kind of 
a consensus-oriented management structure and a style that 
diffused leadership and decision making and didn't take 
responsibility or accountability, and through that, people down 
the line were able to stop progress in its tracks and yet have 
no accountability for that.
    Part of what we have tried to do these last 5 years in 
Congress, as you know, is to improve on that, but my question 
to you is, are those problems still there? Is it the stovepipe 
problem, which is clearly a problem--literally a problem in 
terms of your computer system? It still is, and that is part of 
what we are trying to overcome with the modernization program, 
but there was also this stovepipe sense of----
    Mr. EVERSON. Right.
    Mr. PORTMAN. Having the various units from audit to 
collection to information not working together toward 
modernization. How would you analyze what your management 
problems are in terms of the failures to achieve the goals, 
forgetting again the contractor side of it----
    Mr. EVERSON. Yes.
    Mr. PORTMAN. What would you suggest to address those, or 
have you already addressed them?
    Mr. EVERSON. Let me center my response on these cultural 
observations that you have made, which I think are correct and 
I think, in fact, were perpetuated by the way we have managed 
these projects. We have attempted to achieve consensus in the 
management of these projects and I don't think that is a 
sensible way to approach this. It is important to achieve buy-
in. It is not best practice to achieve consensus when you are 
making difficult decisions like these--the tenets of creating 
these vast systems--because that does give everybody the right 
to circle back and have yet one more change, and the process 
runs 30 or 50 weeks when it should run 5 or 10, and that is 
what happened here.
    I was rather shocked, and as you know, when I got to the 
IRS and we had something like 40 different committees, and 
there was one committee on the modernization and I saw the 
minutes of it and it had 40 people attending the meeting. You 
don't get something done with 40 people trying to discuss a 
computer system. That is not correct, because what you do in 
that model, in my view, is you dumb it down. The guy from GAO 
says, ``Well, we really shouldn't do that.'' The person from 
OMB says, ``Well, we can't do that.'' The person from the U.S. 
Department of the Treasury says, ``We shouldn't do that.'' The 
person from the operating unit says, ``We can't do that.''
    Before you know it, you have taken a lot of the meat out of 
what you need to do. This is all about making tough decisions, 
and standardization, just as you indicated--buying things that 
get the job done at a reasonable cost but may not be the 
Cadillac, when they are available on the outside, and not 
insisting that it needs to be customized and taking forever to 
get it done.
    So, I do think we had a culture that was, as you indicated, 
risk averse. I remember when Todd Grams, who moved over to be 
the Chief Information Officer, he was our Chief Financial 
Officer. He came and briefed me when I was still at OMB and 
said there are no consequences for failure. I would actually 
suggest to you that the action we took yesterday is the first 
time in the course of this most recent program that there has 
been a consequence, a real consequence. So, we are trying to 
change those cultures and those behaviors.
    Mr. PORTMAN. Again, we are going to hear later from the 
contractor and also from the Oversight Board--I see Larry 
Levitan is here--and from your consultant. I guess my only 
thought is from the experience we had going back to the mid-
1990s is that changes in the IRS side, forgetting the 
contractor for a moment, seem to me to be critical if we are 
going to make these targets going forward. I hope you will 
continue to focus on that. I think your announcement this 
morning of the actions you took yesterday is a very positive 
step and the competition is important, but we also need to 
focus on the accountability, as you talked about, and the buy-
in and not the consensus.
    Mr. EVERSON. Absolutely.
    Mr. PORTMAN. Thank you, Mr. Commissioner. Thank you, Mr. 
Chairman.
    Chairman HOUGHTON. Thank you. We have been joined by Mr. 
Weller, and we are going to pass him at the moment because he 
doesn't have a particular question, but Mr. Pomeroy has got 
one. If you think of anything later on, please chime in.
    Mr. POMEROY. Thank you, Mr. Chairman. Lately, we have seen 
an awful lot of what we would have believed to be service work 
of a technical nature performed in this country moved to India 
or other places, and I am wondering if one of the things you 
monitor in the performance of the contract is whether or not 
the work on the IRS is being done within this country.
    Mr. EVERSON. I haven't thought of that issue, and I will 
have to ask that question, frankly. I don't know whether CSC or 
any of the subcontractors that participate with us are doing 
anything overseas. I just don't know the answer to that, and we 
will find that out.
    Mr. POMEROY. I will ask them, as well, but obviously there 
are sensitivities about having our tax system performed on with 
work by contractors out of the country. We would certainly want 
to have this type of project performed in this country, and so 
I would put that on your list of things to keep an eye on. 
Thank you, Mr. Chairman.
    Chairman HOUGHTON. Mr. Portman, have you got any other 
questions?
    Mr. PORTMAN. No.
    Chairman HOUGHTON. Have you got one?
    Mr. WELLER. Mr. Chairman?
    Chairman HOUGHTON. Yes, please.
    Mr. WELLER. Commissioner, thank you for joining us this 
morning. I would echo my friend, Mr. Pomeroy's, concerns. I 
think that is a legitimate question that all taxpayers would 
have if contractors were to send offshore the work. That would 
be a great concern to both parties, and I certainly want to 
echo the question that he asked.
    [Additional information submitted by Mr. Weller follows:]

                                               Washington, DC 20515
                                                  February 25, 2004

The Honorable Mark Everson
Commissioner
Internal Revenue Service
1111 Constitution Avenue, NW
Washington, DC 20224

Dear Mr. Commissioner:

    I am writing you today regarding the Administrations' competitive 
sourcing initiative and the potentially damaging impact it may have on 
my district.
    As part of the initiative, the Internal Revenue Service's Central 
Area Distribution Center in Bloomington is subject to bidding by public 
and private contractors. After contacting your Congressional Affairs 
staff, I learned that a decision will be made on approximately 
Wednesday, April 14, 2004. With over 500 jobs at stake, this is a 
disturbing, potentially damaging development.
    I support the Administrations goal of bringing more efficient, 
lower-cost government to the American people. However, with this many 
jobs at stake in an area that has been hard hit by the downturn in 
employment of the last few years, these jobs will not likely be 
replaced.
    The deciding factor the IRS is using to determine how best to 
administer the Distribution Centers is cost-effectiveness. To that end, 
I offer this resolution. Regardless of whether the Federal Government 
or a contractor receives the right to manage the IRS Distribution 
Centers, I believe the most effective option would be to consolidate 
operations by closing the two facilities in Rancho Cordova, CA and 
Richmond, VA and move those facilities duties to the Bloomington 
location. By closing the two facilities on the coast, you would reduce 
the overall number of employees necessary to operate all three 
facilities, with the added advantage of having a Distribution Center 
centrally located within the Nation, allowing for equal service ability 
nation-wide.
    Thank you for your assistance on this issue. I look forward to 
working with you on finding a positive resolution to this difficult 
situation.

            Sincerely,
                                                       Jerry Weller
                                                 Member of Congress

                               __________

                                               Washington, DC 20515
                                                      March 5, 2004

The Honorable Mark Everson
Commissioner
Internal Revenue Service
1111 Constitution Avenue, N.W.
Washington, DC 20224

Dear Mr. Commissioner:
    My letter serves to update you on my findings after visiting the 
IRS service center in Bloomington, Illinois, and to seek your 
assistance for the center's employees. I am disturbed and disappointed 
to learn that employees have been told that employees must make 
decisions on buyouts before the future of the center has been decided.
    It was brought to my attention that the employees at this facility 
were given a letter and a form on February 29, 2004 for a buyout 
package to return no later than March 20, 2004. While this option was 
taken advantage of by several employees who found it an attractive time 
to leave, there are many more who would rather remain employed at the 
facility, and would prefer to wait until the announcement has been made 
on what will happen to their jobs.
    As the deadline to apply for buyout occurs approximately one month 
prior to the announcement regarding the fate of this facility and the 
people who work there, I ask you to extend the deadline to apply for 
buyout, or offer another buyout opportunity after the IRS makes it's 
announcement. This will allow employees who have submitted buyout 
applications under some duress to rescind them, and reapply later 
should they ultimately decide they would like to be bought out. 
Additionally, for employees who prefer to wait, but are feeling 
pressured to make a decision, this will give them some time and peace 
of mind to make a more fully informed choice.
    I hope you will agree with me that this is a fundamental issue of 
fairness. Please extend the deadline to submit an application for 
buyout, or offer another opportunity for buyout after the IRS announces 
it's decision whether to keep the facility open.
    I look forward to working with you to modernize and streamline the 
IRS while ensuring it's employees are treated fairly.

            Sincerely,
                                                       Jerry Weller
                                                 Member of Congress

                                 

    I think probably the most basic question, which I believe 
just from reviewing what has been discussed this morning, is 
when will the taxpayers begin to see the benefits? When will 
the taxpayers begin to see the results of the computer systems 
modernization? I know in the 9 years that I have served in the 
Congress, we have been talking about this. When are we actually 
going to see the results of the massive investment that 
taxpayers have made in modernizing the computer systems for the 
IRS?
    Mr. EVERSON. I think that it is fair to say that you are 
already seeing the results in many, many dimensions. As I was 
indicating before you came in, we have been successful, and the 
contractor has been successful, in a suite of applications that 
help the taxpayer directly and that help practitioners. So, 
these projects, like e-services, giving Employment 
Identification Numbers, providing transcripts of key data, and 
the history of an account to a practitioner, all those things 
are happening. We have been successful doing that.
    You can see the results in enhanced rate of electronic 
filing. If you look at things like the downloading of forms and 
regulations that are occurring now, it is almost up to a 
billion a year. Think about the savings you are getting there 
because you don't have to call to ask someone for a form and 
then the form goes to one of our distribution centers and gets 
mailed to you. You can just pull it off online. There is a lot 
of progress.
    The problem we have had here, though, is on these major 
systems, the updating of the master file and also the 
infrastructure we will start to use. Our expectation is that 
this the first module of the CADE project, which is the master 
file updating, will actually be working this summer for a 
subsection of 1040EZ filers, several million people. Now, it 
will be a while before you get the follow on modules that 
increasingly pick up other taxpayers. So, this remains, as to 
the big projects, the master file update, a multi-year effort, 
and that will take a lot of time.
    Mr. WELLER. I believe that we have invested about $8 
billion so far in modernization. That is the estimated cost of 
the overall modernization. Is that figure accurate?
    Mr. EVERSON. I presume that in that figure you are going 
back to what Mr. Portman was talking about. If you look at the 
entire life cycle over----
    Mr. WELLER. Right.
    Mr. EVERSON. Back into the early 1990s, a lot of that had 
no benefit whatsoever.
    Mr. WELLER. Yes. Now, you have stated the benefits that you 
have been able to identify that taxpayers are now currently 
receiving. If we were to compare that to a private sector, say, 
a credit card institution or bank or insurance company, how 
today would you rate the level of service as a result of the 
modernization compared to what a customer would receive from 
their bank or financial institution?
    Mr. EVERSON. I think it is improving. I don't think we 
could say that it is quite yet at the level that I would 
consider best practice. A good example would be our telephone 
routing and technology. We have got this large center down in 
Atlanta that routes telephone calls so that if you have got a 
question about charitable contributions, it goes to the right 
person in one of our two dozen or so call centers, and to an 
individual who knows about that subject, and it is also 
sequenced correctly so that you don't have to wait too long 
because of a busy signal.
    I was in a local shop in Arlington just a week ago and a 
fellow who does picture framing for me said, ``You know, I used 
to call the IRS and I used to have to always call on my lunch 
hour because I knew it would take me the full hour to do the 
business. Now when I call, I get right through,'' he said, 
``and unfortunately, at the end of the conversation I have 
usually concluded that you are right and I am wrong.''
    [Laughter.]
    So, I think things are getting better. I wouldn't tell you 
we are best practice yet, though, but we are going to continue 
to work on it.
    Mr. WELLER. Just in follow-up and my last question here, we 
always establish benchmarks----
    Mr. EVERSON. Yes.
    Mr. WELLER. That we measure improvement and measure 
performance. As we look at the coming year, this fiscal year 
that we are currently in, the 2004 fiscal year, between now and 
the end of this year, what benchmark do you believe that we in 
Congress should use to measure the progress as you continue to 
implement and move through the modernization process to 
demonstrate the benefits that we will be able to measure 
progress?
    Mr. EVERSON. In terms of the modernization program itself, 
I think that you need to look at a variety of indicators. Some 
of them are volumes, as I indicated, the percentage of 
electronic filing, or others are more qualitative, like level 
of service. We track level of service, which is a concept that 
does get to this issue of benchmarking against comparable kinds 
of activities where someone is dealing with a reservation 
system, those kinds of issues. So, we have a variety of 
indicators that we track that are summarized, to which we hold 
ourselves accountable.
    The problem that you have got here, that we are really 
talking about this morning, is hitting delivery dates, and as 
the Chairman indicated, controlling costs. I would suggest to 
you that in many instances, when we finally get some of these 
things online, we are getting the functionality that we have 
sought.
    Mr. WELLER. Thank you. Thank you, Mr. Chairman.
    Chairman HOUGHTON. Mr. Portman?
    Mr. PORTMAN. Mr. Chairman, thank you. Just one quick 
follow-up. It seems to me one of the ways to follow on Mr. 
Weller's question, and I appreciate his backing up and saying, 
what is this all about for the taxpayers, because I think that 
we have made some progress. We need to acknowledge that, and 
now we need to get into the underpinnings and the CADE is the 
obvious example there.
    Mr. EVERSON. Yes.
    Mr. PORTMAN. Our CADE is one of the programs that you have 
talked a lot about since coming on last summer, and I think we 
all acknowledge the importance of getting that up and going.
    In your testimony, written testimony, you mention that as a 
milestone. The other one would be the IFS project. Those two 
seem to me ones where you believe that we are close to 
achieving at least significant progress, and by that, I mean 
within the next year. Are those milestones we should hold you 
and the IRS accountable for? Are those programs which, instead 
of talking about the overall BSM, can we talk about those two 
as projects that within the calendar year 2004 we expect to 
make significant progress on?
    Mr. EVERSON. I think that is entirely correct. Let me say 
on the financial system, this is what gave rise to the action 
we took yesterday. We had expected to deliver the first big 
block of that in April. That was a date that had rolled over 
from the beginning of this fiscal year, that is to say October 
1st, and that is now going to be delayed until the end of this 
fiscal year.
    I still believe that there is no reason why that shouldn't 
happen. This is just a SAP financial system. The SAP financial 
systems have been put in complicated corporations across the 
world for a decade or more, so we ought to be able to do this.
    What is difficult in that, and also in the first big piece 
about CADE, is the linkage back into the legacy systems. The 
IRS did a lousy job documenting all the systems changes that it 
made over the many decades. Because nothing was really brought 
current in terms of new investments over those periods of time, 
every time you try to patch back to your human resources system 
or your procurement system, if you are talking about the 
financial piece, or if you try to go back into the code on CADE 
to see what changes were made when the code was revised in 1986 
or in the next year or the next year after that, this has been 
one huge stumbling block that affects both projects. I am sure 
your subsequent panelist, Mr. Cofoni, will address that. It has 
been a really difficult area, in large part because the code 
was never--this is the software code--was never properly 
documented.
    So, that is one issue, and I think that is being addressed, 
and there has been success in testing on the balance and 
control, which was a big issue associated with CADE, and the 
contractor and the IRS now feel better about this first module, 
which is the 1040EZ, coming through.
    The trick will probably not be, I would suggest, achieving 
this first piece, which will only be a narrow strip of 
taxpayers. It will be the follow on work where there is 
something called the business rules engine that will help you 
chart a future path to be able to adjust to all the changes 
that you gentlemen make to the Tax Code and to keep this a 
flexible tool. Setting that up is critical, and you will hear 
about that from one of your later witnesses. That is a whole 
second area that I would say is very significant, about which 
we are concerned. We think we have a good approach here, but we 
are going to need to very much follow that.
    So, the answer is yes, I think both of those deliveries, 
the first piece of CADE and IFS, will happen. We are 
accountable for that. We are accountable with the contractor. 
If neither of those happen, we will take further actions on 
this that would be even more draconian than what was taken 
today.
    Mr. PORTMAN. Thank you. Given what has happened, again, not 
just in the last few years but looking back over the last 
decade, it would seem to me that would make sense, Mr. 
Chairman, for us to focus on a couple of these deliverables, 
including that first module in CADE, within this calendar year 
and even before we go into recess, because it sounds like that 
is about where the fiscal year ends.
    You have a lot of support in this Subcommittee, as you have 
seen this morning, on both sides of the aisle, and it is 
support that is cautiously optimistic, that with your 
leadership, we can get this done. We need some deliverables, 
and we are glad that you have rolled up your sleeves and jumped 
into this with such intensity. We expect and hope you will 
continue that. Thank you, Mr. Chairman.
    Mr. EVERSON. Thank you.
    Chairman HOUGHTON. Help me here. Boil this down. Computer 
systems modernization has got two phases, one, the specific 
things you are trying to do now such as the master files 
update, and then also the quality and the service and the 
documentation and the code changes. What specifically do you 
want to get done this year? Just restate that.
    Mr. EVERSON. I think, as your colleague, Mr. Portman, just 
indicated, the two things that I am really looking to do are 
the same two things that I was looking to do when I was at my 
confirmation hearing in March 2003. They are, unfortunately, a 
year later now. They are the first big block on the financial 
system, IFS, because there will be some follow on work in 
subsequent periods for that and this first piece of CADE, the 
1040EZ filers, that will prove the concept of the new CADE and 
the linkage back into the legacy systems. Those two are the big 
ticket items.
    Chairman HOUGHTON. So, that is what we can look forward to 
quizzing you on next year?
    Mr. EVERSON. That is correct.
    [Laughter.]
    I hope it is not quizzing. I hope that we are all saying, 
well, good, we have turned some corners here, and that with the 
changes you have made and with the improved performance by the 
contractor that we have seen some real deliverables.
    Chairman HOUGHTON. Okay. I have just got one other 
question. The press reports today indicate that the GAO has 
found that 27,000 defense contractors owe about $3 billion in 
back taxes. Have you got any comments on that?
    Mr. EVERSON. In fact, I am heading right over to testify 
before Senators Coleman and Levin on just this issue in just a 
minute. What has been indicated here is that the IRS and the 
U.S. Department of Defense haven't adequately followed up on 
potentially leviable debts that contractors to the government 
have. They haven't paid their taxes but they are still doing 
business with the Department of Defense.
    The President's budget request, which will provide over 
1,000 new people in collections officers, will help. We are 
looking at this issue. I am going to be testifying to the fact 
that we are going to make some short-term procedural changes. 
They are already underway.
    There are statutory concerns here because what you have is 
two competing public policy interests. The first is the 
protection of taxpayer rights, which there are many procedural 
protections written into the law, some through RRA 1998. There 
are also privacy concerns. On the other hand, there is a 
legitimate expectation that if you are doing business with the 
government, you should pay what you owe. Those two intersect, 
and they don't intersect in a positive way. So, there may be 
some statutory changes at which the Congress needs to look.
    Beyond that, I would say we are very anxious to help our 
collection efforts. As you know, in the budget we are asking 
for the ability to take a strip, just a small strip, of the 
moneys due and have private collection agencies help us do 
that, but, as you have heard me testify before, with full 
taxpayer rights. So, I am hopeful we can make some progress in 
this area, Mr. Chairman.
    Chairman HOUGHTON. Mr. Commissioner, thank you very much 
for being here. We appreciate it. Good luck in your further 
testimony.
    Mr. EVERSON. Thank you, gentlemen.
    [Additional questions submitted by Chairman Houghton to Mr. 
Everson, and his responses follow:]

    Question: Many state tax agencies have completed successful 
modernization programs while the IRS has faced cost overruns and 
delays. On the subject of IRS reorganization of its management team 
relating to computer modernization, what are the plans to add people 
and/or contracts that have been part of a prior successful tax 
modernization project?
    Answer: The IRS Business Systems Modernization (BSM) program is as 
complex and challenging as any information technology program in the 
world. When nominated in January 2003, I made it one of my top 
priorities to evaluate the BSM program. I spoke with IRS senior 
executives and other federal government officials as well as outside 
business leaders and executives to learn about the BSM initiative.
    Realizing that the modernization program will only be successful if 
the most senior and experienced IRS executives take ownership of the 
program, I immediately identified the need to strengthen the BSM 
executive management team. I appointed John Dalrymple, a 30-year 
veteran of the IRS who has spent his career focusing on frontline 
taxpayer issues, to become the Deputy Commissioner for Operations 
Support. I also appointed the IRS CFO, Todd Grams, to the CIO position 
to bring better financial and management discipline to the technology 
modernization program.
    I expected the IRS would achieve key milestones in the summer and 
fall of 2003 by delivering the initial release of the Customer Account 
Data Engine (CADE) program in August and the Integrated Financial 
System (IFS) in October. By late summer, it became painfully clear that 
CADE and IFS would experience substantial cost overruns and delays. 
With these two major project setbacks, it was evident that significant 
problems still existed with the BSM program.
    Even before these further delays became clear, the PRIME 
contractor, CSC, and I commissioned three external independent studies 
to assess the health of the modernization program and to review 
specific projects such as the CADE program. Bain and Co., Carnegie 
Mellon Software Engineering Institute (SEI), and GartnerGroup conducted 
these comprehensive assessments during the summer of 2003.
    The IRS developed a 46-item action plan based on feedback from 
these external reviews. The IRS added two more items after reviewing 
recommendations the IRS Oversight Board submitted, bringing the total 
to 48 action items. These recommendations revolved around strengthening 
the BSM executive management team and clearly defining the roles and 
responsibilities between the IRS and the PRIME. As a result, I re-
assigned experienced IRS business leaders to the modernization program. 
I asked Rich Morgante, the Deputy Commissioner for the Tax Exempt and 
Government Entities Division, to join the modernization executive 
management team to oversee the implementation of the recommended action 
items. I also asked John Duder, the Deputy Commissioner, Wage and 
Investment Division, to focus on delivering the CADE project.
    The IRS also retained a leading executive search firm to conduct 
searches for five key executive positions in the modernization program. 
One of these searches is in final negotiations. These external searches 
will provide the IRS with candidates who have a wide-range of diverse 
experience in systems modernization.
    The SEI will also periodically review the CADE program, and we are 
hiring a third party firm to regularly assess the overall health of the 
BSM program. Furthermore, I decided to direct the new enforcement 
projects scheduled for later this year and early next year to other 
contracts.
    All of these measures--combined with the fact that world-class 
firms like IBM, SAP, BearingPoint, Unisys, and Northrop Grumman are 
also engaged in the BSM program--should help us gain better control 
over future cost overruns and project delays.
    Question: I am concerned about taxpayer compliance and debt 
collection. Most financial services companies and state tax agencies 
have implemented modern compliance systems to ensure delinquent debts 
are collected. You have placed an increased emphasis on enforcement and 
proposed a new Filing & Payment Compliance (F&PC) initiative that would 
bring in billions of dollars of additional revenue each year. What is 
currently being done to ensure that this project will move forward, 
while the Business Systems Modernization initiative is being 
reexamined?
    Answer: The BSM initiative is a key part of our broader agenda at 
the IRS. Our focus is service + enforcement = compliance. While the IRS 
is continually improving service to make it easier for taxpayers to 
understand and comply with the tax laws, it is simultaneously boosting 
its enforcement initiatives.
    Seeing as billions and billions of dollars are lost in tax revenue 
each year, President Bush's FY 2005 budget submission requests an 
additional $300 million over the FY 2004 consolidated appropriations 
level for enforcement activities. The IRS has begun to address this tax 
gap crisis by shifting badly needed resources and hiring more frontline 
enforcement personnel, who will focus primarily on noncompliance among 
high income individuals and businesses.
    The Filing & Payment Compliance (F&PC) project is funded to start 
in FY 2005. The first steps in implementing the new F&PC series of 
projects--which we often call Collection Contract Support--involve 
using private collection agencies. It will provide support to enable 
private collection agencies to supplement the IRS's internal collection 
staff; however, using private collection agencies to resolve delinquent 
taxpayer cases requires enabling legislation (for which I ask your 
support).

                                 

    Chairman HOUGHTON. We are now going to have a panel 
consisting of Larry Levitan, who is a Member of the IRS 
Oversight Board; Steve Palmquist, Chief Engineer for Civil and 
Intelligence Agencies at Carnegie Mellon University in 
Arlington; Robert Dacey, Director of the Accounting and 
Information Management Division of the GAO; and Paul Cofoni, 
President of the Federal Sector and Corporate Vice President of 
CSC.
    Larry, good to see you again, and would you commence with 
your testimony.

  STATEMENT OF THE HONORABLE LARRY LEVITAN, MEMBER, INTERNAL 
                REVENUE SERVICE OVERSIGHT BOARD

    Mr. LEVITAN. Good morning. Mr. Chairman, Members of the 
Subcommittee, my message today is somber, so I will dispense 
with any formalities and get right to my testimony.
    During this past summer, the modernization program fell 
into a ditch, experiencing significant delays and budget 
overruns in virtually all of the projects underway. Although we 
are convinced that the overall modernization plan is sound and 
well designed, the challenge is executing that plan.
    Unfortunately, this is not the first ditch that the 
modernization program has fallen into. Since its inception 5 
years ago, it has had, with few exceptions, a consistent track 
record of missed target dates and budget estimates. All seven 
major application projects the IRS has undertaken are both 
significantly over cost and behind schedule.
    Without making excuses, the modernization program is 
extremely large and complex with numerous risks. Programs of 
this scale, and there are very few, are never completed without 
some level of cost and schedule overruns. It is also important 
to recognize that the modernization program has had a number of 
important successes as described by the Commissioner.
    It should also be realized that Commissioner Everson and 
his executive teams are new to their responsibilities and to 
direction of the modernization program. As soon as the overruns 
described previously this summer became apparent, the 
Commissioner started a comprehensive review of modernization 
and has now initiated an aggressive program to address the 
problems.
    The Board has been impressed with the speed, thoroughness, 
and openness of this effort. The program of improvement 
identified by this effort has been started, and we believe that 
meaningful progress is being made.
    The problems experienced by the modernization program will 
not be solved easily or, in many cases, quickly. However, they 
can be solved. In its recently released report on the program, 
the Board makes nine specific recommendations.
    Number one, business unit management must take direct 
leadership and ownership of the modernization program, and in 
particular this must include defining the scope of each 
project, preparing realistic and attainable business cases, and 
controlling scope changes.
    Number two, create an environment of trust, confidence, and 
teamwork between the business units, the BSM and Information 
Technology Services organizations, and the PRIME.
    Number three, enhance the systems development life cycle 
methodology to support more accurate estimating of future work 
phases and put into place the necessary processes to ensure 
that the methodology is followed religiously.
    Number four, enhance the contracting process.
    Number five, significantly strengthen the experience and 
capabilities of the IRS's BSM team. The capabilities of this 
team must be consistent with the scale and complexity of the 
modernization program.
    Number six, rationalize the oversight of the program to 
eliminate duplication and streamline the process.
    Number seven, the CADE project requires special attention. 
The CADE is the foundation of modernization. It is also by far 
the most costly, complex, largest, and longest running project 
within modernization. Additional work is necessary to ensure 
that the development approach selected for this system will 
work effectively.
    Number eight, reduce the number of projects being conducted 
concurrently. Modernization currently has five major projects 
underway. As demonstrated by the program's performance, neither 
the IRS nor the PRIME has the capability to manage and conduct 
this many projects at the same time. It is the Board's view 
that as current work phases are completed, the number of 
projects should be reduced. As both the IRS and PRIME 
demonstrate they can strengthen their capabilities and improve 
their performance, the portfolio of concurrent projects can be 
increased.
    Number nine, the IRS Oversight Board has lost confidence in 
the leadership of the PRIME contracting team. This conclusion 
has been reached after observing performance and results for 
over 3 years. There have been numerous commitments to enhance 
the capabilities of this team, improve its management 
processes, and deliver greater thought and management 
leadership. Overall results have not changed, however. Target 
dates and budgets are consistently missed.
    In fact, Mr. Chairman, just last week, the IRS was informed 
that the target date for IFS would have to be changed once 
again. In our view, this was the final straw. The Board now 
strongly recommends that the IRS consider all options for 
strengthening the capabilities of the PRIME contracting team. 
The time for strong, aggressive action is now.
    Mr. Chairman, our judgments may be harsh but are made 
because of our firm belief that the modernization program 
cannot be allowed to fail. The risk to the country is 
unacceptable. The IRS cannot continue to operate with the 
outmoded and inefficient systems and processes it uses today. 
Over time, the existing systems will become impossible to 
maintain, and at that point, the ability to administer our 
country's tax system is at risk.
    We are convinced that the overall plan is sound and well 
designed. The challenge is executing that plan. Meaningful 
recommendations have been made by each of the groups involved 
in this recent reassessment. The Commissioner and his 
management team are committed to the improvement program and 
the IRS Oversight Board supports that plan and will continue to 
work with the IRS on this important effort.
    [The prepared statement of Mr. Levitan follows:]

  Statement of the Honorable Larry Levitan, Member, Internal Revenue 
                        Service Oversight Board

Introduction and Summary
    Mr. Chairman, and members of the Subcommittee, thank you for 
holding this hearing and inviting me to testify. It is an honor for me 
to appear before your committee today on behalf of the IRS Oversight 
Board. My remarks today will be directed at the IRS' efforts to 
modernize its computer systems.
    The long-term health and viability of the nation's tax 
administration system rest upon the success of the IRS Business Systems 
Modernization program (BSM or Modernization). However, during this past 
summer, the BSM program suffered a serious setback. Virtually all of 
its ongoing projects experienced significant delays and budget 
overruns.
    Particularly troubling were continuing and unresolved problems with 
the Customer Account Data Engine (CADE)--the so-called ``crown jewel'' 
of Modernization--that will move taxpayers from the current antiquated 
tape-based system to a modern reliable data base.
    To his credit, IRS Commissioner Mark W. Everson quickly called for 
a number of separate independent reviews of not only CADE, but the 
entire BSM portfolio of projects. While the assessments' results are 
still preliminary, they make it clear that the IRS and its Prime 
contractor cannot continue to operate in a business-as-usual manner. 
The IRS Oversight Board believes that the stakes are too high and BSM's 
problems are too severe to be addressed with half-measures. They must 
be squarely addressed in rigorous and open fashion, and as soon as 
possible.
    Based on the findings and recommendations of the BSM reviews, 
Commissioner Everson recently launched an aggressive ``action plan'' to 
remedy the problems plaguing Modernization. Through this report, the 
Board makes nine specific recommendations for turning around the 
critical BSM program.
    They include having the IRS business units take greater leadership 
and ownership of BSM projects, enhancing the contracting processes, 
improving overall program management and focus, and reducing the number 
of BSM projects underway at the same time.
    Of great significance, the Board also strongly recommends that the 
Prime's performance be closely monitored and, if significant 
improvements are not quickly demonstrated, a change should be made.
    On balance, the IRS Oversight Board is convinced that the overall 
Modernization plan is sound and well-designed. No one believes that the 
IRS should start over from scratch; a firm foundation and architecture 
are in place. However, the challenge for the IRS and the Prime remains 
how to execute that plan and successfully implement the new systems and 
processes on schedule and within budget.
A Troubled History
    The BSM program is crucial to delivering better service to 
taxpayers and increasing the Agency's efficiency and productivity. Over 
the past few years, some notable BSM projects and benefits were 
delivered. Indeed, tangible improvements in call routing, e-Filing and 
interactive services produced enhanced service for taxpayers and more 
efficient operations at the IRS.
    However, last summer, the BSM program appeared at the point of 
unraveling. Virtually all of the projects with a major impact on 
improving customer service and IRS' internal operations and 
productivity were experiencing serious delays and cost overruns.

      The Customer Account Data Engine (CADE), which will 
replace the IRS Master Files with a modern database management system 
and provide the foundation for other modernized applications, missed 
its Release 1 August ``go-live'' date. As of this report's publication, 
a new target date was not finalized.
      Very little work was done to confirm that the ``business 
rules engine,'' on which future CADE releases depend, will be able to 
handle the complexity and scale of the CADE data base.
      The October ``go-live'' target date for the Integrated 
Financial System (IFS), which will eventually replace IRS' old core 
financial systems, was missed. Release 1 was rescheduled for the spring 
of 2004 and work on Release 2 was deferred. Costs increased 
proportionately.
      Modernized e-File, the platform for all internet tax 
return forms, was also delayed, increasing project cost.
      The Custodial Accounting Project (CAP) project, which 
will provide the IRS with critical control and reporting capabilities, 
was also running significantly behind schedule and over budget.
      E-Services, which provides a suite of web-based products 
to third-party users, is basically on schedule, but significantly over 
budget.

    Unfortunately, this was not the first time Modernization found 
itself in serious trouble. Cost overruns have become an all-too 
familiar story. Since its inception, BSM had, with few exceptions, a 
consistent track record of missed target dates and budget estimates. 
For example, the CADE project is now over two years behind schedule and 
has been re-scheduled four times. The only reason it is only $30 
million over the original cost estimate is that the IRS converted CADE 
to a fixed-price contract shortly after the project first started to 
experience problems.
    CADE is not the only BSM project that has a track record of 
failure. This is a shared BSM problem. Overall, the e-Services project 
is 28 months behind schedule and $72 million over the original budget, 
due to both scope growth and cost increases. The IFS project is almost 
$50 million over the original cost estimate and although it is only two 
months behind schedule at present, there is no current firm estimate of 
when it will be delivered. In fact, all seven major application 
projects the IRS has undertaken are both over cost and behind schedule.
Getting Behind the Problem
    Why did this happen? How did BSM end up with a consistent track 
record of overruns? Without making excuses, the Modernization program 
is extremely large and complex with numerous risks. Programs of this 
scale, and there are very few, are never completed without some level 
of cost and schedule overruns. However, looking back with 20/20 
hindsight, BSM's problems are all too evident and severe. The following 
are some of the more prominent deficiencies.
    There was inadequate business unit ownership and sponsorship of 
projects. This resulted in unrealistic business cases and continuous 
project scope ``creep''.
    The much desired environment of trust, confidence and teamwork 
between the IRS business units, the BSM organization, the Information 
Technology Services (ITS) organization, and the Prime simply did not 
exist. In fact, the opposite was true, resulting in an inefficient 
working environment and, at times, finger pointing when problems arose.

      The project life cycle methodology did not fully support 
the requirement to estimate future phases of the work. Moreover, the 
methodology was not always rigorously followed. At times, inappropriate 
shortcuts were taken in order to meet unrealistic target dates, further 
exacerbating problems.
      The contracting process was highly inefficient. This 
caused significant extra overhead for both the Prime and IRS and at 
times, resulted in work being done without a contract or inappropriate 
contracts being used.
      The BSM organization did not have the depth and breadth 
of skills and experience to adequately manage the Modernization program 
and the Prime contract.
      As the program ran into more and more problems, 
additional layers of review and auditing were put into place by OMB, 
the Treasury Inspector General for Tax Administration (TIGTA), GAO, and 
Congress. While this may be understandable, given BSM's history, it 
added significant overhead to program management.
      Computer Sciences Corporation (CSC), the Prime team's 
leader, did not demonstrate that it had the depth of leadership and 
experience to successfully carry out its responsibilities. The Prime 
team's track record was marred by continuous delays, missed target 
dates and budget overruns. CSC did not supply the important thought and 
program leadership it was engaged to deliver. Up until the last few 
months, CSC was unable to develop a strong working relationship with 
the IRS' executive leadership.
      Based on all of the above factors, program productivity 
levels were extremely low.
      The program schedule was too ambitious given the 
capabilities of both the IRS and the Prime. There were too many 
concurrent projects. More became less.

    It should be understood that Commissioner Everson and his executive 
team are new to their responsibilities and to the direction of the 
Modernization program. However, as soon as delays and overruns 
experienced this summer became apparent, the Commissioner launched a 
comprehensive review of Modernization.
    The Board was impressed with the speed, thoroughness and openness 
of this endeavor. It should also be noted that CSC participated fully 
with this effort and in fact, hired an independent organization to 
assess its own work. The program of improvements identified by the 
review process recently began and while a number of difficult decisions 
must still be made, the Board believes that meaningful progress is 
being made. Commissioner Everson is to be commended for his swift 
response. However, more must be done.
The Board's Recommendations
    The problems outlined in this report do not yield to quick or easy 
solutions. However, this does not mean the IRS is confronting a 
hopeless situation. BSM's problems can be managed and solved. To this 
end, the Board makes the following nine recommendations which it 
believes will help set BSM back on the path to success. In most cases, 
they are consistent with those made by the independent experts engaged 
by the IRS and the Prime. In some cases, action has already started.
    Recommendation 1: The IRS business units must take direct 
leadership and ownership of the Modernization program and each of its 
projects. In particular, this must include defining the scope of each 
project, preparing realistic and attainable business cases and 
controlling scope changes throughout each project's life cycle. The 
Deputy Commissioner for Operations Support assumed responsibility for 
this critical task and has already taken steps to insure that it is put 
into place.
    Recommendation 2: Create an environment of trust, confidence and 
teamwork between the business units, the BSM and ITS organizations, and 
the Prime. This is a cultural issue and will take time. The Deputy 
Commissioner for Operations Support and the CIO have responsibility for 
this action. CSC is also actively involved.
    Recommendation 3: Enhance the systems development life cycle 
methodology to support more accurate estimates of future work phases 
and put into place the necessary processes to insure that the 
methodology is followed religiously. Again, this work is under way.
    Recommendation 4: Enhance the program's contracting process and 
capabilities.
    Recommendation 5: Significantly strengthen the experience and 
capabilities of the BSM team. Its capabilities must be consistent with 
the scale and complexity of the Modernization program. This will 
require a meaningful number of outside hires from organizations, such 
as the Defense Department, that have experience with large, complex 
programs.
    Recommendation 6: Try to rationalize the oversight of the program 
to streamline the process and eliminate duplication. This will require 
a joint effort of the IRS, OMB, TIGTA, GAO, the Oversight Board and 
Congress.
    Recommendation 7: The CADE project requires special attention. CADE 
is the foundation of Modernization. It is also by far the most costly, 
complex, largest and longest running project within the BSM portfolio. 
As part of the review program initiated by the Commissioner, the 
Software Engineering Institute (SEI) was engaged to review CADE, with 
special emphasis on the systems architecture, conversion approach and 
the planned use of a ``business rules engine'' to make the development 
and maintenance of the system more efficient.
    SEI's findings were generally supportive of the system design and 
the use of the business rules engine. However, SEI could not be sure 
that the engine could handle the size and complexity of the CADE system 
until the rules were defined and modeled. It recommended that a project 
to define and model the rules--a major effort--be conducted as soon as 
possible. The Board strongly supports this recommendation.
    Recommendation 8: Reduce the number of projects being conducted at 
the same time. Modernization currently has five major projects 
underway. As demonstrated by the BSM program's performance, neither the 
IRS nor the Prime has the capability to manage and conduct this many 
projects at the same time.
    In addition, defining and modeling CADE's business rules (as 
recommended above) will be a significant additional effort. The Board 
believes that as current work phases are completed, the number of 
projects should be reduced. This must happen carefully and over time, 
so that current work efforts are not lost.
    Admittedly, this will be a very difficult decision for the IRS to 
make. Each of the Modernization projects is important and brings 
significant business value. However, the Board strongly believes that 
such a step is absolutely necessary for the overall success of 
Modernization. As both the IRS and Prime demonstrate that they can 
strengthen their capabilities and improve their performance, the 
portfolio of concurrent projects can be increased.
    Recommendation 9: The IRS Oversight Board has lost confidence in 
the leadership of the Prime contracting team. This conclusion has been 
reached after observing performance and results for over three years. 
There have been numerous commitments to enhance the capability of this 
team, improve its management processes and deliver greater thought and 
management leadership. Overall results have not changed, target dates 
and budgets are consistently missed. In fact Mr. Chairman, just last 
week the IRS was informed that the target date for IFS would have to be 
changed once again. In our view this was the final straw. The Board now 
strongly recommends that the IRS consider all options for strengthening 
the capabilities of the Prime contracting team. The time for strong, 
aggressive action is now.
    The IRS Oversight Board now strongly recommends that the IRS take 
meaningful action to correct this problem. The IRS needs to consider 
alternative actions to correct this problem.
Conclusion
    The IRS Oversight Board firmly believes that the IRS Modernization 
program cannot be allowed to fail. The IRS cannot continue to operate 
with the outmoded and inefficient systems and processes it uses today. 
Over time, the existing systems will become impossible to maintain and 
at that point, the ability to administer our country's tax system will 
be in grave danger. Such a risk to our nation is unacceptable. We 
remain convinced that the overall Modernization plan is sound and well-
designed. The challenge is executing that plan. The IRS and the Prime 
must get it right this time.
    Meaningful recommendations were made by each of the groups involved 
in this recent reassessment to improve the management and execution of 
Modernization. In this same vein and spirit, the Board adds its 
constructive suggestions today.
    The Commissioner and his management team are committed to a 
Modernization improvement program. The Oversight Board supports that 
plan and will continue to work with the IRS to make sure that this 
important effort succeeds.

                                 

    Chairman HOUGHTON. Thank you, Mr. Levitan. Mr. Palmquist?

STATEMENT OF M. STEVEN PALMQUIST, CHIEF ENGINEER FOR CIVIL AND 
 INTELLIGENCE AGENCIES, ACQUISITION SUPPORT PROGRAM, SOFTWARE 
ENGINEERING INSTITUTE, CARNEGIE MELLON UNIVERSITY, PITTSBURGH, 
                          PENNSYLVANIA

    Mr. PALMQUIST. Mr. Chairman and Members of the 
Subcommittee, thank you for inviting me here today. My name is 
Steven Palmquist. I work for Carnegie Mellon University's 
Software Engineering Institute (SEI), a federally funded 
research and development center with a mission to help 
government and industry improve their software engineering and 
acquisition practices.
    At the SEI, I am the Chief Engineer for Civil and 
Intelligence Agencies within our Acquisition Support Program. 
Today, I represent a team of 14 principal and senior members of 
our staff who performed an Independent Technical Assessment 
(ITA) of the IRS' CADE program during the late summer and early 
fall of 2003.
    At the SEI, we have performed over 50 ITAs, usually at the 
request of senior Department of Defense and civil agency 
leaders when they have a program in trouble. For CADE, 
Commissioner Everson was our senior sponsor and we appreciated 
his candor and support.
    In each of the ITAs we have performed, we have been 
impressed with the dedication, intelligence, and 
resourcefulness of the people involved on both the government 
and the industry side. The CADE was no exception. This then 
forces a basic question. With all of these good, talented 
people, why is CADE failing?
    The answer is because CADE, unfortunately, is again not an 
exception. The CADE fell into the same traps that have crippled 
many other government and industry programs, both large and 
small. The CADE's program planning has been deficient, its 
execution has been uncoordinated. Measurements and metrics have 
not been properly utilized, and risks have not been identified, 
profiled, and managed.
    In its simplest form, project management has two functions, 
plan the work and work the plan. This did not happen. 
Requirements in particular were not fully understood or 
communicated, and history has shown that budgets and schedules 
were unrealistic.
    We believe the CADE team needs to adopt a ``back to 
basics'' approach. First, they need to restore rigor and 
discipline in their management processes, both programmatic and 
technical. They must take a long-term total systems view of 
CADE through a solid systems engineering effort. They must 
begin the expensive but critical task of harvesting the 
business rules. They need to strengthen and support their 
current software development environment, but continue to 
investigate improvements to that environment, such as a 
business rules engine. Last, they should continue to seek 
outside reviews of their program, a commercial best practice, 
and one of the key recommendations of the November 2000 report 
of the Defense Science Board on Defense Software.
    The CADE by its nature is high risk. We believe the CADE 
team can deliver. We believe the talent and the desire is 
there. It will not happen, however, until the IRS and its 
industry partners improve their management discipline. We did 
see evidence of improvements during our ITA. However, many of 
these processes are still not effectively executed and many are 
not backed by sufficient technical expertise and experience.
    In particular, we have significant concerns with the lack 
of emphasis on the harvesting of the business rules. There was 
also not enough information available to determine if the 
architecture is viable in Releases 3, 4, and 5, and we are 
especially concerned security is not adequately addressed as 
CADE moves from its current once a week batch processing 
environment to the more real-time interactive approach of 
Releases 3, 4, and 5.
    Mr. Chairman, thank you again for the opportunity to be 
here today. The SEI is proud of the work we did in supporting 
Commissioner Everson and the IRS, and I look forward to 
answering any questions you may have.
    [The prepared statement of Mr. Palmquist follows:]

    Statement of M. Steven Palmquist, Chief Engineer For Civil and 
     Intelligence Agencies, Acquisition Support Program, Software 
    Engineering Institute, Carnegie Mellon University, Pittsburgh, 
                              Pennsylvania

Introduction
    Mr. Chairman and members of the Subcommittee, thank you for 
inviting me to speak to you here today. My name is Steven Palmquist. I 
work for Carnegie Mellon University's Software Engineering Institute. I 
am the Chief Engineer for civil and intelligence agencies in our 
Acquisition Support Program. My background includes licensure as a 
Professional Engineer and certification as a Project Management 
Professional.
    Today I represent a team of fourteen principal and senior members 
of our technical staff who performed an independent technical 
assessment of the IRS's Customer Account Data Engine (CADE) program. 
This work was done in the late summer/early fall of 2003.

What is the Software Engineering Institute?
    Founded in 1984, the Software Engineering Institute, or SEI as it 
is more-commonly known, is a federally funded research and development 
center (FFRDC). Sponsored by the U.S. Department of Defense, the SEI's 
mission is to help DoD agencies, other government agencies and industry 
improve their software engineering and acquisition capabilities.

What is an Independent Technical Assessment (ITA)?
    An Independent Technical Assessment, or ITA, is an objective 
examination of a project or program conducted by outside experts. As an 
FFRDC located at a major university, the SEI is fiercely protective of 
its credibility and neutrality in conducting these types of 
assessments. Because of our objectivity, senior military and civil 
agency executives often request we assess their software-intensive 
programs. To do this, we form a team with the appropriate mix of 
expertise drawn from across the Institute. We gather information 
through interviews, fact-finding visits to facilities, reviews of 
designs and software architectures, data reviews, and process reviews. 
We analyze this information and present our findings and 
recommendations to the senior executive sponsor.
    For CADE, IRS Commissioner Everson was our senior executive 
sponsor.

Background of the CADE Program and the Critical Role of Business Rules
    In his May 20, 2003 testimony to the Annual IRS Restructuring and 
Reform Act of 1998 Joint Congressional Review, Commissioner Everson 
stated, ``the centerpiece of the modernization effort is the 
replacement of the decades-old Master Files . . . (with) a modern, 
reliable database, called the Customer Account Data Engine, or CADE.'' 
IRS modernization thus depends on CADE. CADE, as we discovered, depends 
on business rules.
    The IRS and their industry partners agreed on a business rules 
approach for CADE in a contract awarded in December 1998. Simply put, a 
business rule is any statement that defines how a business conducts its 
business. For the IRS, business rules are principally representations 
of tax laws and tax forms. The potential benefit of a business rules 
system is the separation of the business logic from program logic. For 
example, a business rules approach would allow the IRS to easily change 
the rules to reflect new tax year changes without affecting the 
underlying computing environment.
    CADE's original timetable called for five yearly software 
increments or releases. Release 1, scheduled for January 2002, was 
primarily the infrastructure needed for all CADE releases. It also 
contained the approximately 1,200 business rules needed to process 
1040EZ filers. Subsequent CADE releases significantly increase both the 
number and complexity of business rules such that by Release5, all of 
CADE's estimated 50,000 business rules are included. With Release 5, 
the IRS would process nearly all individual tax filers using CADE.
    The Release 1 business rules were originally to be coded using a 
business rules engine, a commercial software tool. A business rules 
engine interprets a formalized representation of the natural-language 
expression of a business rule to generate the software code that 
implements the business rule. A rules engine offers the promise of 
faster rules coding, which will be critical in CADE's later releases, 
where the huge majority of rules reside. In fact, all of CADE's future 
cost, schedule, and performance goals depend on the assumed efficiency 
of a rules engine.
    However, the Release 1 business rules engine effort stalled over 
contractual issues. The IRS's industry partners then coded the business 
rules using the C++ programming language. This switch was one of the 
principal reasons Release 1 missed the original January 2002 delivery 
date, which was rescheduled for August 2003. However, the plan remained 
to use a business rules engine on subsequent releases, including a 
revised Release 1, so that all CADE business rules would still be 
implemented and managed by the rules engine.

Background of the SEI's ITA of CADE
    Because of the continued focus on a business rules engine, the IRS 
requested we perform a ``health check'' on its use on CADE. On July 1, 
2003, we began this work.
    On July 25, 2003, however, Commissioner Everson announced that CADE 
Release 1 would also miss the new August 2003 delivery date. Coincident 
with this announcement, Commissioner Everson expanded the scope of the 
health check to a full ITA of all of CADE.
    We completed the ITA and presented our report to Commissioner 
Everson in October 2003. In the four months since, both the IRS and 
their industry partners have worked on our findings and 
recommendations. We have not been involved in these efforts, so we have 
not updated our report. The following findings and recommendations, 
therefore, represent CADE as it stood in the late summer and early fall 
of 2003.

Principal Findings
Delivery of Release 1
    We believe the IRS and its industry partners can deliver CADE 
Release 1--with the functionality/design as described to us during the 
ITA--by August or September of 2004. This is provided no new 
requirements emerge and the IRS and their industry partners agree to 
the acceptance criteria. As expressed to Commissioner Everson, our 
confidence in this date was approximately sixty percent.

A Business Rules Approach
    We agree that a business rules approach offers the IRS a 
potentially significant capability to manage and improve their 
operations. Reaching that potential with a system as large as CADE, 
however, will require a more-disciplined approach and execution by both 
the IRS and their industry partners. As an example, at the time of our 
ITA there was no definitive evidence that the rules engine would 
perform adequately in the IRS's operational environment or in their 
industry partners' development, test, and maintenance environment.

Long-term Vision
    We believe the current emphasis on Release 1 has been at the 
expense of a long-term perspective. Any complex software system 
involves balancing competing requirements of performance, security, 
interoperability, maintainability, modifiability, etc.--commonly called 
the quality attributes. For CADE, those requirements are implemented in 
a sequence of coordinated steps in successive releases of the 
modernized system. The choices of the step content, and the balancing 
of these critical system properties, depend on how the IRS will use the 
modernized system at each stage. However, this analysis has not been 
adequately addressed for future releases. This is particularly true for 
security concerns as the IRS moves from its current ``lock down'' 
security environment into the interactive environment of later CADE 
releases.

Software Capability
    While a number of CADE's problems can be attributed lapses in the 
industry partners' software development processes, we believe the 
industry partners can deliver quality software. Recent improvements in 
systems engineering, measurement and metrics, and technical reviews 
have resulted in improved performance. However, several key development 
processes are still not effectively integrated, and there is no clear 
chain of command for technical decision-making.

Management Capability
    We saw evidence that both the IRS and their industry partners are 
moving toward more effective management discipline. One example of this 
is the adoption of the Software Acquisition Capability Maturity Model 
(SA-CMM). However, not all their processes are effectively executed, 
and many are not backed by sufficient technical expertise and 
experience. Risk management, requirements management, staffing and 
talent retention, communications management and creating accurate 
budgets and schedules all remain areas of concern.

Principal Recommendations
Harvest the IRS Business Rules
    We believe that harvesting the business rules, not coding them, 
will drive the cost and schedule of future CADE releases. By 
harvesting, we mean capturing, adjudicating, and cataloging the rules. 
CADE has invested many resources exploring rules engines, but few 
resources exploring the rules themselves. The IRS needs to understand 
and document their business rules as well as the rules' complicated 
interactions. Some of the delays that have already plagued CADE are a 
direct result of an imperfect understanding of the business rules. This 
situation will only grow as the number and complexity of the 
implemented rules increases.
    In addition, the IRS does not know the number of business rules in 
CADE with any reasonable degree of certainty. Therefore, no one knows 
how long rule harvesting will take, how many people will be required, 
the background, training and experience of the people required, or how 
much it will cost. Based on anecdotal information presented to us, we 
believe the time will be measured in years and cost will be measured in 
the tens of millions of dollars.
    Until sound, supported cost and schedule estimates for rule 
harvesting are available, future CADE plans and schedules are only 
tentative and likely subject to delays and missed milestones.

Institutionalize Systems Engineering
    CADE needs a strong systems engineering approach to identify and 
handle programmatic and technical risks. To do this, the IRS's industry 
partners need to establish a permanent CADE systems engineering group. 
We recommend the IRS and its industry partners use ``model problems'' 
to help understand the risks and potentially uncover new, critical 
unknowns.
    The IRS also needs dedicated system engineers to staff the systems 
engineering group. These persons would be responsible for understanding 
CADE's technical issues, as well as the industry partners' proposed 
solutions, for their impact on the IRS. In particular, the IRS should 
lead an integrated team to define, monitor, manage, and support 
operational security. This effort should be across not only CADE but 
also the entire modernization effort and the IRS's legacy computing 
environment.

Validate Business Rules Engines
    As stated earlier, all of CADE business rules are currently coded 
in C++. While this was not the original plan, it is the only design 
approach for which CADE has historical data, and the only approach 
validated in the CADE environment. Therefore, C++ is the baseline 
design approach and should be the basis for future planning.
    However, other design approaches--the rules engine, improving the 
C++ environment, or other programming options--should be examined. If 
they have merit, they should be rigorously validated in the CADE 
environment. If any of these alternatives prove superior(in terms of 
performance, risk, cost, schedule, etc.), they could then become the 
primary design approach. Coding the rules in C++ then becomes the 
fallback design approach, but one with known cost, schedule, and 
performance data.

Institutionalize Management Discipline
    The IRS and its industry partners must both improve their 
management discipline. While the following list is in no way complete, 
we believe both parties should:

      continue to build a collaborative, active management 
structure
      establish real, supported risk management programs to 
aggressively identify and manage risk
      improve their ability to jointly create realistic cost 
and schedule estimates
      coordinate testing activities
      create a baseline set of known requirements, and 
characterize the ``known unknowns'' for all releases
      determine the communication needs of all CADE 
stakeholders, and lay out a plan for how needed information will be 
distributed, received and used
      establish mechanisms to independently monitor these 
activities
      capture lessons learned and feed the information forward 
to future activities

Plan for Independent Oversight
    Because large, complex, software-intensive system acquisitions face 
a number of challenges, they often can benefit from experienced and 
trusted software expertise. For example, the Army's Future Combat 
System (FCS) program has a software steering committee to help identify 
issues and risks and offer an overall ``sanity check'' one step removed 
from day-to-day operations.
    In the FCS model, both the government and contractor call on the 
members of the software steering committee. The group functions to 
advise, and not to critique, senior management. The group is a partner 
rather than another oversight body. Recent lessons learned from the 
first phase of the FCS program have cited the software steering 
committee as an indispensable factor in achieving program goals.
    A second type of oversight is an independent expert review, similar 
to our ITA. Independent expert reviews are an industry best practice 
and one of the key recommendations of the November 2000 Report of the 
Defense Science Board Task Force On Defense Software, issued by the 
Office of the Under Secretary of Defense For Acquisition and 
Technology.
    Independent expert reviews last one-to-two days and are held two to 
three times a year. The reviews are designed to help program teams 
ensure:

      disciplined processes and methodologies are in place
      the program is adequately resourced
      the technical baseline is understood and solid, with 
attendant risks and opportunities identified and managed
      adequate progress is being achieved

    To be effective, the independent expert reviews need to be part of 
CADE, with these two defining elements:

      the reviews are scheduled parts of, not intrusions into, 
CADE's management and development plans, schedules and processes
      CADE's program and technical management are robust and 
flexible enough to pursue the opportunities and address the weaknesses 
uncovered by the reviews

    Without these two conditions, the value of the reviews diminishes 
exponentially.

Conclusion
    We commend the IRS and their industry partners for seeking review 
from outside sources. Incorporating insight and recommendations from 
other organizations (such as commercial financial organizations, other 
government agencies, academia, etc.) will yield a richer set of 
recommendations to help put CADE back on track. However, any 
recommendation--including our own--must be regularly evaluated for 
appropriateness in the CADE environment, as well as for their benefit 
to the IRS in terms of cost, schedule, and performance.
    Mr. Chairman, thank you again for the opportunity to be here today. 
The SEI is proud of the work we did supporting Commissioner Everson and 
the IRS. I look forward to answering any questions you may have.

                                 

    Chairman HOUGHTON. Thanks very much, Mr. Palmquist. Mr. 
Dacey?

 STATEMENT OF ROBERT F. DACEY, DIRECTOR, INFORMATION SECURITY 
             ISSUES, U.S. GENERAL ACCOUNTING OFFICE

    Mr. DACEY. Mr. Chairman and Members of the Subcommittee, I 
am pleased to be here today to discuss IRS' system 
modernization efforts. I will briefly summarize my written 
statement.
    The IRS' attempts to modernize its aging computer system 
span many years. A history of continuing delays and design 
difficulties led to GAO designating IRS's system modernization 
efforts as high-risk in 1995. The IRS' current effort, BSM, was 
initiated in fiscal year 1999. To date, about $1.7 billion has 
been appropriated, including about $388 million for fiscal year 
2004.
    To facilitate Congressional oversight of this program, 
annual appropriations laws have mandated that modernization 
funds not be available until IRS submits to the Appropriations 
Committees for approval a modernization expenditure plan that 
satisfies certain legislative conditions, including a review by 
GAO. We are currently reviewing the fiscal year 2004 
expenditure plan.
    In prior reviews of these plans, we have identified 
numerous deficiencies in the BSM program and provided 
recommendations to address them, most importantly, balancing 
the pace of systems acquisition projects with the agency's 
ability to manage them, and also establishing repeatable 
processes for acquiring software and improving modernization 
management controls and capabilities, such as those related to 
configuration management and cost and schedule estimating.
    In response to our recommendations, IRS has made important 
progress which the Commissioner discussed earlier this morning. 
Nevertheless, as we reported last June, IRS continued to face 
challenges to fully develop and implement its management 
capabilities.
    Our written testimony provides an analysis of the reported 
cost overruns and schedule delays that have affected most of 
the current BSM projects. In addition to the deficiencies I had 
previously discussed, our work has shown that the increases and 
delays were caused in part by several factors, including 
inadequate definitions of system requirements, increases in 
project scope, and underestimating project complexity. These 
schedule delays and cost overruns have impaired IRS' ability to 
make appropriate decisions about investing in projects, have 
delayed the delivery of benefits to taxpayers, and postponed 
the resolution of material weaknesses in other IRS program 
areas.
    Given the continued cost overruns and schedule delays, IRS 
and CSC launched internal and independent assessments during 
2003 on the health of BSM as a whole and CADE in particular. 
These more in-depth and comprehensive assessments provided an 
analysis of BSM weaknesses and risks consistent with our prior 
findings that contributed to these delays. The assessments also 
provided actionable recommendations to address the weaknesses.
    Based on these assessments, IRS developed action plans for 
46 specific issues that it identified for resolution, 27 of 
which they have reported were completed at the end of last 
month. Also, IRS has contracted with MITRE to conduct an 
independent analysis of the efficacy of these action plans.
    Significant further work remains to complete implementation 
of the remaining 19 issues. The IRS is also taking other 
action, such as planning to have SEI conduct further periodic 
reviews of the CADE project. Additionally, IRS is responding to 
recommendations from the Oversight Board, which Mr. Levitan 
summarized briefly a moment ago, and from the Treasury 
Inspector General for Tax Administration. The IRS has reported 
they expect to fully implement remaining open actions by the 
end of this calendar year.
    Commitment of appropriate resources and top management 
attention are critical to meeting these challenges and 
improving BSM performance. In addition, continuing oversight by 
Congress, OMB, and others, as well as ongoing assessments of 
the program, can assist IRS in strengthening the program.
    Mr. Chairman and Members of the Subcommittee, this 
concludes my statement. I will be pleased to answer any 
questions that you have.
    [The prepared statement of Mr. Dacey follows:]

 Statement of Robert F. Dacey, Director, Information Security Issues, 
                     U.S. General Accounting Office

Mr. Chairman and Members of the Subcommittee:

    I am pleased to be here today to discuss the Internal Revenue 
Service's (IRS) actions to modernize its computer systems. Although 
updated through the years, IRS's set of computer systems is based on an 
architecture that dates from the 1960s. This architecture has inhibited 
IRS's ability to effectively and efficiently perform its mission of 
providing service to taxpayers and enforcing the nation's tax laws. 
However, IRS's attempts to modernize its computer systems and 
underlying architecture now span three decades. Given the long history 
of continuing delays and design difficulties, we previously designated 
IRS's modernization program as a high-risk area in 1995.\1\ It remains 
so today.\2\
---------------------------------------------------------------------------
    \1\ U.S. General Accounting Office, High-Risk Series: An Overview, 
GAO/HR-95-1 (Washington, D.C.: February 1995).
    \2\ U.S. General Accounting Office, High-Risk Series: An Update, 
GAO-03-119 (Washington, D.C.: January 2003).
---------------------------------------------------------------------------
    IRS's current multibillion-dollar effort, known as the Business 
Systems Modernization (BSM) program, was initiated in fiscal year1999. 
IRS contracted with Computer Sciences Corporation (CSC) as the prime 
contractor to assist with designing, developing, and integrating a new 
set of information systems that were intended to replace IRS's aging 
business and tax processing systems. To date, about $1.7 billion has 
been appropriated for the program, including about $388 million for 
fiscal year 2004.\3\
---------------------------------------------------------------------------
    \3\ P.L. 108-199, Div. F, Title II, Jan. 23, 2004. IRS uses the 
appropriated totals to cover contractor costs related to the BSM 
program. IRS funds internal costs for managing BSM with another 
appropriation. These costs are not tracked separately for BSM-related 
activities.
---------------------------------------------------------------------------
    To facilitate congressional oversight of this program, annual 
appropriations laws since fiscal year 1998 have mandated that 
modernization funds not be available until IRS submits to the 
congressional appropriations committees for approval a modernization 
expenditure plan that satisfies six legislative conditions, including 
that it be reviewed by us.\4\ We are currently reviewing the fiscal 
year 2004 BSM expenditure plan. During our past reviews of such plans, 
we have noted numerous modernization management control deficiencies 
and made recommendations to correct them. Although IRS has made 
progress in implementing our recommendations, BSM continues to face 
significant challenges and serious risks. Recognizing these risks, IRS 
and CSC recently completed several in-depth and more comprehensive 
assessments on the health of the BSM program, including an independent 
technical assessment of the Customer Account Data Engine (CADE) 
project, a project critical to the success of BSM. IRS has developed an 
action plan to address the assessments' recommendations, and has begun 
to act on it.
---------------------------------------------------------------------------
    \4\ The other five legislative conditions are that the expenditure 
plan (1) meets Office of Management and Budget's (OMB) capital planning 
and investment control review requirements; (2) complies with IRS's 
enterprise architecture; (3) conforms with IRS's enterprise life cycle 
methodology; (4) is approved by IRS, Treasury, and OMB; and (5) 
complies with federal acquisition rules, requirements, guidelines, and 
system acquisition management practices.
---------------------------------------------------------------------------
    In my testimony today I will summarize our prior findings and 
recommendations and those of the recently completed program 
assessments. I will also discuss the actions IRS reports it has taken 
or plans to take to address issues raised by these assessments.
    In preparing this testimony, we relied on our prior reports and 
testimony on IRS's systems modernization activities and BSM expenditure 
plans. We also reviewed and analyzed information contained in the BSM 
expenditure plan for fiscal year 2004; Carnegie Mellon University 
Software Engineering Institute's (SEI) independent technical assessment 
of CADE; reports on the BSM program by the Treasury Inspector General 
for Tax Administration and the IRS Oversight Board; and IRS briefing 
materials (1) analyzing the root causes of BSM project cost increases 
and schedule delays, (2) independent reviews of CSC's business 
processes and IRS's procurement practices, and (3) IRS's action plan to 
address issues identified by the reviews. We did not independently 
validate planned projects' cost estimates or confirm, through system 
and project management documentation, the validity of IRS-provided 
information on the projects' content and progress. Our work was 
performed during the past month, in accordance with generally accepted 
government auditing standards.

Background
    The tax administration system that collects about $2 trillion in 
revenues each year is critically dependent on a collection of obsolete 
computer systems developed by the IRS over the last 40 years. IRS 
envisions a future in which its tax processing environment will be 
virtually paper-free, and up-to-date taxpayer information will be 
readily available to IRS employees to respond to taxpayer inquiries. To 
accomplish this, IRS embarked on its ambitious BSM program. BSM 
involves the development and delivery of a number of modernized 
business, data, and core infrastructure projects that are intended to 
provide improved and expanded service to taxpayers as well as IRS 
internal business efficiencies. Recognizing the long-term commitment 
needed to solve the problem of obsolete computer systems, Congress set 
up a special BSM account in fiscal year 1998 to fund IRS's systems 
modernization efforts.
    IRS initiated CADE as part of BSM, to modernize the agency's 
outdated and inefficient data management system.\5\ IRS also sees this 
project as the corporate data source enabling future customer service 
and financial management applications. CADE is therefore IRS's linchpin 
modernization project. In light of the projects that depend on CADE, as 
well as the many interrelationships that are to exist among CADE and 
IRS's modernized applications and among CADE and current IRS 
applications, the agency must manage this critical project effectively. 
Without CADE, the business systems modernization program cannot 
succeed.
---------------------------------------------------------------------------
    \5\ The current system--referred to by IRS as the master files--
contains taxpayer account and return data. There are master files for 
individuals, businesses, and employer retirement plans. A nonmaster 
file for taxpayer data also exists that cannot be stored in the other 
master files due to data format and space limitations.
---------------------------------------------------------------------------
IRS Has Made Improvements, But Systems Modernization Program Remains 
        High-Risk
    IRS's attempts to modernize its aging computer systems span several 
decades. This long history of continuing delays and design difficulties 
led to our designating IRS's Tax Systems Modernization program, BSM's 
predecessor, as a high-risk area in 1995.\6\ During the mid-1990s we 
reported on several technical and management weaknesses associated with 
Tax Systems Modernization, a program that began in the 1980s. These 
weaknesses related to incomplete or inadequate strategic information 
management practices; immature software development capability; 
incomplete systems architecture, integration planning, system testing, 
and test planning practices; and the lack of an effective 
organizational structure to consistently manage and control systems 
modernization organizationwide. We made a series of recommendations for 
correcting these weaknesses and limiting modernization activities until 
they were corrected.\7\ IRS subsequently discontinued the program after 
the agency had spent about $4 billion without receiving expected 
benefits. In fiscal year 1999, IRS launched the BSM program. IRS 
contracted with CSC as its prime systems integration services 
contractor for systems modernization, helping it design new systems and 
identify other contractors to develop software and perform other tasks.
---------------------------------------------------------------------------
    \6\ GAO/HR-95-1.
    \7\ U.S. General Accounting Office, Tax Systems Modernization: 
Management and Technical Weaknesses Must Be Corrected If Modernization 
Is to Succeed, GAO/AIMD-95-156 (Washington, D.C.: July 26, 1995) and 
Tax Systems Modernization: Blueprint Is a Good Start, But Not Yet 
Sufficiently Complete to Build or Acquire Systems, GAO/AIMD/GGD-98-54 
(Washington, D.C.: Feb. 24, 1998).
---------------------------------------------------------------------------
    In our reviews of IRS's BSM expenditure plans, we have identified 
numerous deficiencies in the BSM program, including a continuation of 
the weaknesses noted above. Also, a consistent challenge for IRS has 
been to make sure that the pace of systems acquisition projects does 
not exceed the agency's ability to manage them. In May and November 
2000, we reported that projects were in fact getting ahead of the 
modernization management capacity that needed to be in place to manage 
them effectively.\8\ In February 2002 we reported that such an 
imbalance was due to IRS's first priority and emphasis being on getting 
the newer, more modern systems--with their anticipated benefits to 
taxpayers--up and running.\9\ In so doing, however, management controls 
had not been given equal attention and thus had not kept pace. This 
emphasis on new systems added significant cost, schedule, and 
performance risks that escalate as a program advances. Moreover, these 
risks increased as IRS moved forward because of interdependencies among 
projects, and the complexity of associated workload activities to be 
performed increased dramatically as more systems projects were built 
and deployed.
---------------------------------------------------------------------------
    \8\ U.S. General Accounting Office, Tax Systems Modernization: 
Results of Review of IRS' March 7, 2000, Expenditure Plan, GAO/AIMD-00-
175 (Washington, D.C.: May 24, 2000) and Tax Systems Modernization: 
Results of Review of IRS' August 2000 Interim Spending Plan, GAO-01-91 
(Washington, D.C.: Nov. 8, 2000).
    \9\ U.S. General Accounting Office, Business Systems Modernization: 
IRS Needs to Better Balance Management Capacity with Systems 
Acquisition Workload, GAO-02-356 (Washington, D.C.: Feb. 28, 2002).
---------------------------------------------------------------------------
    In addition, we identified other deficiencies in the BSM program, 
including the need to establish processes that meet the level 2 
requirements of the SEI's Software Acquisition Capability Maturity 
ModelTM,\10\ and to improve modernization management 
controls and capabilities, such as those related to configuration 
management, risk management, enterprise architecture implementation, 
human capital strategic management, integrated program scheduling, and 
cost and schedule estimating.
---------------------------------------------------------------------------
    \10\ Carnegie Mellon University's Software Engineering Institute 
has developed criteria, known as the Software Acquisition Capability 
Maturity ModelTM, for determining organizations' software 
acquisition management effectiveness or maturity. A Level 2 
organization has established its basic project management processes in 
the following key process areas: acquisition planning, solicitation, 
requirements development and management, project management, contract 
tracking and oversight, evaluation, and transition to support.
---------------------------------------------------------------------------
    In response to our recommendations, IRS has made important 
progress. First, significant progress has been made in establishing the 
modernization management controls needed to effectively acquire and 
implement information technology systems. For example, IRS has

      invested incrementally in its modernization projects;
      defined a systems life cycle management methodology, 
which IRS refers to as the Enterprise Life Cycle;
      developed and is using a modernization blueprint, 
commonly called an enterprise architecture, to guide and constrain its 
modernization projects; and
      established processes that meet the level 2 requirements 
of the SEI's Software Acquisition Capability Maturity 
ModelTM.

    Second, IRS has made progress in establishing the infrastructure 
systems on which future business applications will run. For example, 
IRS has delivered elements of the Security and Technology 
Infrastructure Release to provide the hardware, software, and security 
solutions for modernization projects. IRS has also built an enterprise 
integration and test environment that provides the environment and 
tools for multiple vendors associated with a release to perform 
integration and testing activities.
    Third, it has delivered certain business applications that are 
producing benefits today. These applications include

      Customer Communications 2001, to improve telephone call 
management, call routing, and customer self-service applications;
      Customer Relationship Management Examination, to provide 
off-the-shelf software to IRS revenue agents to allow them to 
accurately compute complex corporate transactions; and
      Internet Refund/Fact of Filing, to improve customer self-
service by providing to taxpayers via the Internet instant refund 
status information and instructions for resolving refund problems.

    Fourth, IRS took steps to align the pace of the program with the 
maturity of IRS's controls and management capacity, including 
reassessing its portfolio of planned projects.
    Nevertheless, IRS continued to face challenges to fully develop and 
implement its modernization management capacity. Last June we reported 
that IRS had not yet fully implemented a strategic approach to ensuring 
that it has sufficient human capital resources for implementing BSM, 
nor had it fully implemented management controls in such areas as 
configuration management, estimating costs and schedules, and employing 
performance-based contracting methods.\11\ We made several 
recommendations to address those issues. Our analysis has shown that 
weak management controls contributed directly to the cost, schedule, 
and/or performance shortfalls experienced by most projects. Given that 
the tasks associated with those projects that are moving beyond design 
and into development are by their nature more complex and risky and 
that IRS's fiscal year 2004 BSM expenditure plan supports progress 
toward the later phases of key projects and continued development of 
other projects, systems modernization projects likely will encounter 
additional cost and schedule shortfalls. IRS will need to continue to 
assess the balance between the pace of the program and the agency's 
ability to manage it.
---------------------------------------------------------------------------
    \11\ U.S. General Accounting Office, Business Systems 
Modernization: IRS Has Made Significant Progress in Improving Its 
Management Controls, but Risks Remain, GAO-03-768 (Washington, D.C.: 
June 27, 2003)
---------------------------------------------------------------------------
Projects Continue to Incur Cost Increases and Schedule Delays
    Based on IRS's expenditure plans, BSM projects have consistently 
cost more and taken longer to complete than originally estimated. Table 
1 shows the life cycle variance in cost and schedule estimates for 
completed and ongoing BSM projects. These variances are based on a 
comparison of IRS's initial and revised cost and schedule estimates to 
complete initial operation \12\ or full deployment \13\ of the 
projects.
---------------------------------------------------------------------------
    \12\ Initial operation refers to the point at which a project is 
authorized to begin enterprisewide deployment.
    \13\ Full deployment refers to the point at which enterprisewide 
deployment has been completed and a project is transitioned to 
operations and support.

         Table 1: IRS BSM Project Cost/Schedule Variance Summary
------------------------------------------------------------------------
                                  Reported/
                        Cost       revised   Schedule   Reported/revised
      Project         variance    estimated  variance      estimated
                         (in      cost (in      (in     completion date
                     thousands)  thousands)   months)
------------------------------------------------------------------------
Completed Projects
------------------------------------------------------------------------
Security and
 Technology
 Infrastructure
 Release 1
------------------------+$7,553-----$41,287--------+5------------1/31/02
                                                                (initial
                                                              operation)
------------------------------------------------------------------------
Customer                 +5,310      46,420        +9            2/26/02
 Communications                                        (full deployment)
 2001
------------------------------------------------------------------------
Customer                 -1,938       7,375        +3            9/30/02
 Relationship                                          (full deployment)
 Management Exam
------------------------------------------------------------------------
Human Resources            +200      10,200         0           12/31/02
 Connect Release 1                                              (initial
                                                              operation)
------------------------------------------------------------------------
Internet Refund/        +12,923      26,432       +14            9/26/03
 Fact of Filing                                        (full deployment)
------------------------------------------------------------------------
Ongoing Projects a
------------------------------------------------------------------------
Modernized e-File       +17,057      46,303      +4.5            3/31/04
 Release 1                                                      (initial
                                                              operation)
------------------------------------------------------------------------
e-Services              +86,236     130,281       +18            4/30/05
                                                       (full deployment)
------------------------------------------------------------------------
CADE Release 1          +36,760      97,905      +30b          6/30/05 b
                                                       (full deployment)
------------------------------------------------------------------------
Integrated              +53,916     153,786     TBD b              TBD b
 Financial System                                      (full deployment)
 Release 1
------------------------------------------------------------------------
Custodial               +72,058     119,219     TBD b              TBD b
 Accounting Project                                    (full deployment)
 Release 1
------------------------------------------------------------------------
Customer Account          TBD c       TBD c     TBD c             TBD c
 Management Release
 1
------------------------------------------------------------------------
Source: GAO analysis of data contained in IRS's BSM expenditure plans.
a Projects ongoing as of 9/30/03.
b Project schedules for CADE, the Integrated Financial System, and the
  Custodial Accounting Project are currently under review.
c To be determined. Work on the Customer Account Management project was
  suspended following the completion of preliminary design activities.
  No further work is planned until at least fiscal year 2005.

    As the table indicates, the cost and schedule estimates for full 
deployment of the e-Services project have increased by just over $86 
million and 18 months, respectively. In addition, the estimated cost 
for the full deployment of CADE release 1 has increased by almost $37 
million, and project completion has been delayed by 30 months. In 
addition to the modernization management control deficiencies discussed 
above, our work has shown that the increases and delays were caused, in 
part, by

      inadequate definitions of systems requirements. As a 
result, additional requirements have been incorporated into ongoing 
projects.
      increases in project scope. For example, the e-Services 
project has changed significantly since the original design. The scope 
was broadened by IRS to provide additional benefits to internal and 
external customers.
      cost and schedule estimating deficiencies. IRS has lacked 
the capability to effectively develop reliable cost and schedule 
estimates.
      underestimating project complexity. This factor has 
contributed directly to the significant delays in the CADE release 1 
schedule.
      competing demands of projects for test facilities. 
Testing infrastructure capacity is insufficient to accommodate multiple 
projects when testing schedules overlap.
      project interdependencies. Delays with one project have 
had a cascading effect and have caused delays in related projects.

    These schedule delays and cost overruns impair IRS's ability to 
make appropriate decisions about investing in new projects, delay 
delivery of benefits to taxpayers, and postpone resolution of material 
weaknesses affecting other program areas.
    Producing reliable estimates of expected costs and schedules is 
essential to determining a project's cost-effectiveness. In addition, 
it is critical for budgeting, management, and oversight. Without this 
information, the likelihood of poor investment decisions is increased.
    Schedule slippages delay the provision of modernized systems' 
direct benefits to the public. For example, slippages in CADE will 
delay IRS's ability to provide faster refunds and respond to taxpayer 
inquiries on a timely basis.
    Delays in the delivery of modernized systems also affect the 
remediation of material internal management weaknesses. For example, 
IRS has reported a material weakness associated with the design of the 
master files. CADE is to build the modernized database foundation that 
will replace the master files. Continuing schedule delays will place 
resolution of this material weakness further out into the future. In 
addition, the Custodial Accounting Project is intended to address a 
financial material weakness and permit the tracking from submission to 
disbursement of all revenues received from individual taxpayers. This 
release has yet to be implemented, and a revised schedule has not yet 
been determined. Finally, the Integrated Financial System is intended 
to address financial management weaknesses. When IRS submitted its 
fiscal year 2003 BSM expenditure plan, release 1 of the Integrated 
Financial System was scheduled for delivery on October 1, 2003. 
However, it has yet to be implemented, and additional cost increases 
are expected.

Internal and Independent Assessments of BSM Have Identified Significant 
        Weaknesses and Risks
    Given the continued cost overruns and schedule delays experienced 
by these BSM projects, IRS and CSC launched internal and independent 
assessments during 2003 of the health of BSM as whole, as well as CADE. 
Table 2 describes these assessments.

             Table 2: BSM Assessments Undertaken During 2003
------------------------------------------------------------------------
                         Organization conducting
        Subject                 assessment                Purpose
------------------------------------------------------------------------
Root cause analysis      IRS                      To review data from
                                                   historical documents
                                                   andinterviews to
                                                   determine root causes
                                                   for schedule delays
                                                   and cost increases
------------------------------------------------------------------------
PRIME review             Bain and Company         To identify root
                                                   causes of breakdown
                                                   in CSC'sbusiness
                                                   processes and
                                                   engagement model and
                                                   provide recommended
                                                   solutions
------------------------------------------------------------------------
IRS Office of            Acquisition Solutions,   To assess the
 Procurement Assessment   Inc.                     efficiency and
                                                   effectiveness of the
                                                   IRSprocurement
                                                   organization
                                                   structure, employment
                                                   of best practices,
                                                   managementand
                                                   administration,
                                                   staffing, and to
                                                   briefly review BSM
                                                   contracting
------------------------------------------------------------------------
CADE assessment          SEI                      To provide an
                                                   independent technical
                                                   assessment ofCADE
                                                   program history and
                                                   the feasibility of
                                                   future plans
------------------------------------------------------------------------
Source: IRS

    The IRS root cause analysis, PRIME review, and the Office of 
Procurement assessment revealed several significant weaknesses that 
have driven project cost overruns and schedule delays, and also 
provided a number of actionable recommendations for IRS and CSC to 
address the identified weaknesses and reduce the risk to BSM. 
Deficiencies identified are consistent with our prior findings and 
include

      poorly defined requirements,
      low program productivity levels,
      project scope creep,
      IRS/PRIME role confusion,
      immature management processes,
      ineffective integration across IRS, and
      insufficient applications and technology engineering.

    As noted, CADE release 1 has experienced significant reported cost 
overruns and schedule delays throughout its life cycle, and has yet to 
be delivered. SEI's independent technical assessment of CADE pointed to 
four primary factors that have caused the project to get off track and 
resulted in such severe cost and schedule impairments: (1) the 
complexity of CADE release 1 was not fully understood; (2) the initial 
business rules engine effort stalled; (3) both IRS and PRIME technical 
and program management were ineffective in key areas, including 
significant breakdowns in developing and managing CADE requirements; 
and (4) the initially contentious relationship between IRS and PRIME 
hindered communications. SEI also warned that CADE runs the risk of 
further trouble with later releases due to unexplored/unknown 
requirements; security and privacy issues that have not been properly 
evaluated (e.g., online transactions are different from the way IRS 
does business today); dependence on an unproven business rules engine 
\14\ software product; and the critical, expensive, and lengthy 
business rules harvesting \15\ effort that has not yet been started. 
SEI offered several recommendations to address current CADE issues and 
reduce project risk in the future.
---------------------------------------------------------------------------
    \14\ A business rules engine translates business rules, or 
processing criteria (e.g., income tax refunds of $x or more are held 
for administrative review), into executable computer code which 
processes transactions related to a tax form, and selects and executes 
correct rules based on the tax year and tax form.
    \15\ Business rules harvesting refers to the process of extracting, 
defining, and documenting tax processing criteria from a variety of 
sources, including IRS subject matter experts, legacy system source 
code, the tax code, and various other paper documents.
---------------------------------------------------------------------------
IRS Is Acting to Resolve Issues Identified in the BSM Assessments
    Based on these assessments, IRS identified a total of 46 specific 
issues for resolution in the following six areas, and developed a BSM 
action plan comprising individual action plans to address each issue:

      Organization and Roles. Immediate steps are needed to 
clarify IRS/PRIME roles and responsibilities and clearly define 
decision-making authorities.
      Key Skills & Strengthening the Team. Strengthened skills 
and capabilities are needed in such key areas as project management and 
systems engineering.
      Technology--Architecture & Engineering. More focus is 
needed to improve current systems architecture integration.
      Technology--Software Development Productivity & Quality. 
Improvements in product quality and productivity are essential to 
strengthening software delivery performance.
      Acquisition. Contracting and procurement practices 
require major streamlining to improve overall contract management.
      CADE. Delivery of CADE release 1 will require aggressive 
focus and attention, and a business rules engine solution requires 
additional evaluation.

    These 46 issue action plans were assigned completion dates and an 
IRS or PRIME owner was assigned to take the lead in implementing each 
plan. IRS and PRIME each also assigned a senior-level executive to 
drive the execution of the issue action plans, identify and help 
mitigate implementation hindrances or roadblocks, and ensure successful 
completion of all planned actions. To assess the efficacy of the BSM 
action plan, MITRE was tasked with conducting an independent analysis 
and provided feedback to IRS on the effectiveness of the specific issue 
action plans to address the associated findings/recommendations and 
correct any problems found.
    IRS has reported making steady progress with implementing the BSM 
action plan. According to the IRS BSM program office, as of late 
January 2004, 27 of the 46 issue action plans have been completed. 
Examples of completed actions include (1) making business owners and 
program directors accountable for project success; (2) assigning teams 
to investigate and resolve problem areas on key projects such as CADE, 
the Integrated Financial System, and e-Services; (3) aligning critical 
engineering talent to the most critical projects; (4) increasing the 
frequency of CADE program reviews; and (5) issuing a firm fixed-price 
contracting policy.
    Significant further work remains to complete implementation of the 
remaining 19 open issue action tasks. Bain & Company--which conducted 
the independent review of PRIME--has been hired to facilitate the 
implementation of various issue action plans within the Organization 
and Roles challenge area, while IRS has also contracted with SEI to 
conduct further periodic reviews of the CADE project.
    Additionally, the IRS Oversight Board recently issued a report \16\ 
on its own independent analysis of the BSM program, which made several 
observations and recommendations that are consistent with those 
discussed here. IRS has conducted an analysis of this report to 
reconcile the board's recommendations with those that are currently 
being addressed in the BSM action plan. As a result, IRS plans to open 
two additional issues and action plans to address (1) rationalizing and 
streamlining oversight of the BSM program, and(2) determining and 
maintaining a manageable portfolio of projects. IRS expects to complete 
the majority of the BSM action plan by end of April of this year, and 
fully implement any remaining open actions by the end of the calendar 
year.
---------------------------------------------------------------------------
    \16\ IRS Oversight Board Special Report, Independent Analysis of 
IRS Business Systems Modernization, December 2003.
---------------------------------------------------------------------------
    Further, during 2003, the Treasury Inspector General for Tax 
Administration performed several reviews related to management of the 
BSM program and for specific BSM projects. These reviews identified 
several issues, including those related to compliance with the defined 
management and project development processes, full implementation of 
disciplined project testing processes and procedures, IRS's cost and 
schedule estimation process, and contract management. IRS management 
reaffirmed their commitment to fully implement key management and 
project development processes.

Concluding Observations
    IRS's multibillion-dollar BSM program is critical to agency's 
successful transformation of its manual, paper-intensive business 
operations and fulfilling its restructuring activities. The agency has 
made important progress in establishing long-overdue modernization 
management capabilities and in acquiring foundational system 
infrastructure and some applications that have benefited the agency and 
the public. However, our reviews, those of the Treasury inspector 
general, and the recently completed internal and independent 
assessments of the BSM program clearly demonstrate that significant 
challenges and serious risks remain. IRS acknowledges this and is 
acting to address them.
    To successfully address these challenges and risks and to modernize 
its systems, IRS needs to continue to strengthen BSM program management 
by continuing efforts to

      balance the scope and pace of the program with the 
agency's capacity to handle the workload, and
      institutionalize the management processes and controls 
necessary to resolve the deficiencies identified by the reviews and 
assessments.

    Commitment of appropriate resources and top management attention 
are critical to resolving the identified deficiencies. In addition, 
continuing oversight by the Congress, OMB, and others, as well as 
ongoing independent assessments of the program, can assist IRS in 
strengthening the BSM program.
    Meeting these challenges and improving performance are essential if 
IRS and the PRIME contractor are to successfully deliver the BSM 
program and ensure that BSM does not suffer the same fate as previous 
IRS modernization efforts.
    Mr. Chairman, this concludes my statement. I would be pleased to 
respond to any questions that you or other members of the subcommittee 
may have at this time.

                                 

    Chairman HOUGHTON. Thank you very much. Well, Mr. Cofoni, 
you are on the hot seat.
    Mr. COFONI. Yes, I am, sir.
    [Laughter.]
    Chairman HOUGHTON. Delighted to have you here.
    Mr. COFONI. Thank you.
    Chairman HOUGHTON. Please give your testimony.

    STATEMENT OF PAUL COFONI, PRESIDENT FEDERAL SECTOR, AND 
  CORPORATE VICE PRESIDENT, COMPUTER SCIENCES CORPORATION, EL 
                      SEGUNDO, CALIFORNIA

    Mr. COFONI. Mr. Chairman, Mr. Pomeroy, Mr. Portman, I 
welcome this opportunity to testify today. I am Paul Cofoni, 
President of the Federal Sector and Corporate Vice President of 
CSC.
    Since December 1998, we have led the PRIME Alliance for the 
IRS. The PRIME Alliance includes some of the best names in the 
technology and business modernization world, SAIC, IBM, Unisys, 
Northrop-Grumman, and BearingPoint. In addition to our alliance 
partners, there are about 809 small business subcontractors 
performing work on the program, all of which, by the way, are 
doing that work in this country and principally in Maryland and 
New Carrollton, across the street from the IRS large building 
complex.
    While there have been difficulties with the BSM program, in 
partnership with the IRS, we have created, in fact, a strong 
program with a sound architecture and a technology foundation 
for future success. The IRS's past difficulties in modernizing 
its technology are well documented. Past attempts lacked the 
central vision and architecture, and most of all, failed to 
achieve business objectives that would benefit taxpayers or 
provide significant efficiencies to the government.
    In 1998, after 2 years of competitive bidding, the IRS 
awarded the CSC PRIME Alliance team a contract for up to 15 
years with an original estimated value of up to $7 billion. 
Under the contract, work is identified by task order and 
separately procured, competitively separately procured.
    Through December 2003, Congress has appropriated 
approximately $1.35 billion, and the IRS has funded CSC for 
approximately $927 million of that $1.35 billion. That has been 
done through 114 task orders and approximately 1,100 task order 
modifications.
    Since beginning the program, the PRIME Alliance and the IRS 
have delivered key program and technology foundational elements 
and business applications for the BSM program. These include an 
enterprise architecture and transition strategy that has been 
awarded best in class this past year within government; an 
enterprise life cycle process tailored to the IRS's specific 
needs that creates a methodology for implementing the 
enterprise architecture; and a software acquisition model that 
has been evaluated by Carnegie Mellon's SEI at Level 3. I would 
add, CSC is the first company in the world to receive Level 3 
accreditation for software acquisition. We also assisted the 
IRS in achieving a Level 2, and the IRS is the first civilian 
agency to receive this accreditation.
    We delivered a secure technology infrastructure that allows 
citizens and tax professionals to interact with the IRS in real 
time and to conduct day-to-day business and solve tax problems 
on the Internet. The infrastructure is in place today. It is 
stable, it is secure, and it will serve as the cornerstone for 
successful deliveries of improved services in the future.
    An enterprise management system was also delivered that 
allows IRS information technology professionals the ability to 
monitor this new infrastructure in real time and make 
corrections or preempt problems in real time.
    We also implemented a bilingual customer communication 
technology upgrade which doubles the capacity of the IRS's 
ability to handle telephone calls at its call centers. It 
reduces wait times and helps taxpayers who are seeking 
assistance, as the Commissioner pointed out in his testimony.
    We implemented the Internet Refund Fact of Filing, or 
``Where is my refund?'' web-based application which provides 
taxpayers instantaneous status of a refund versus driving a 
large volume of inquiries into the phone system, causing delay 
and frustration for taxpayers. I would add here that the IRS 
website in the last week in July was 1 of the top 10 websites 
in terms of popularity or hits, and it was primarily this 
application that did that.
    Added to this application in a very short 2-month period 
was the Advanced Child Tax Credit Initiative, which allows 
taxpayers to ascertain whether their tax credit is available to 
them. Taken together, these two applications since 
implementation have avoided over 24 million taxpayers from 
having to make a call to the IRS call center.
    We implemented a laptop software tool that allows IRS 
agents to use modernized and sophisticated technology for the 
first time to do their work. This has resulted in faster 
resolution of cases and consistent treatment across all 
business taxpayers.
    We implemented the Internet Employer Identification Number. 
This allowed over 450,000 applications to be processed over the 
Internet for new small business owners. As the Commissioner 
pointed out, we implemented seven e-services, web-based 
applications that have allowed 55,000 applications for Tax 
Identification Numbers (TIN) by tax preparers, 40,000 
electronic return originators, 40,000 of them to register with 
the IRS, and in the first 24-hour period of an application for 
bulk TIN matching, over 425,000 transactions were completed.
    This being said, there is more to do. We have near-term 
deliveries. The IFS, we have talked about here. We were 
enormously disappointed that we had to miss our commitment 
which we made to the Commissioner. We promised April and we 
will not be able to do that. The reason is that during final 
testing of this application, we encountered data conditions 
that were unknown to ourselves and to the IRS in the legacy 
systems. This makes--these data conditions, which had not been 
defined in our requirements or in the system design, require us 
to go back and do redesign and rebuild and retest the system, 
which will delay implementation until the end of the government 
fiscal year.
    While I am not satisfied with having to delay IFS, we 
understand what caused these problems and we understand what we 
need to do to prevent them going forward. Additionally, I would 
add that CSC has stepped forward and offered to incur all the 
additional costs between April and October for IFS.
    Modernized e-file is a contract that is outside of our 
PRIME contract but for which we do have responsibility for 
integrating that application onto the secure infrastructure and 
CADE Release 1. This is the all-important first step toward 
replacing the master file. We have had great recent progress 
and recently we have passed system integration testing, IRS 
user acceptance testing, and we have conducted a pilot for the 
1040EZ release. We have added 2003 tax law changes to that 
system. Those changes are now undergoing testing, and we are 
planning any necessary 2004 changes so that we will be ready 
for the 2005 filing system to do a complete processing of 
1040EZ.
    Notwithstanding these accomplishments and imminent 
accomplishments and progress, we are not satisfied with our 
performance nor that of our alliance partners. Together with 
the Commissioner, we conducted several studies, as has been 
mentioned. One of those studies we commissioned ourselves and 
we asked Bain and Co. to perform that study to look at where we 
are having difficulties, and here is what Bain concluded.
    First, that the IRS and CSC need to improve significantly 
the business requirements definition process. As the 
Commissioner pointed out, we can no longer allow new business 
requirements to be identified during the testing phases of a 
project. They must be identified at the front end of a project.
    Second, both parties need to streamline the decisionmaking 
process. We must have a single source of decision making 
throughout the program. Consensus decisionmaking among many is 
not feasible for a program of this complexity.
    Third, there must be an increased focus on business 
transformation, more business representation on project teams, 
and an increased role by the business in the requirements 
definition process and transition planning.
    Last, we need to improve the accountability for ourselves 
and our PRIME Alliance partners. We have to hold ourselves 
accountable for our performance, and we do.
    The CSC from the very top of our organization is committed 
to program success, and we will do whatever it takes to 
deliver. The SEI study stated, ``Stay the course.'' I cannot 
agree more. While challenges lie ahead, I know the CSC team in 
place today can face those challenges. The IRS and CSC must 
leverage our past lessons learned and take advantage of our 
successes to improve the overall delivery, performance, and 
eliminate on-time, on-budget issues that persist.
    The goal of CSC as the integrator for this important 
program is to deliver the best tax administration system to the 
American taxpayers. I am personally committed to achieve this 
goal in partnership with the IRS, Congress, and the various 
stakeholders. Thank you.
    [The prepared statement of Mr. Cofoni follows:]
Statement of Paul Cofoni, President Federal Sector, and Corporate Vice 
    President, Computer Sciences Corporation, El Segundo, California
    Mr. Chairman, thank you for this opportunity to testify before the 
Ways and Means Subcommittee on Oversight. Over the past few years, the 
management and employees of Computer Science Corporation have 
appreciated the opportunity to work with the members of this committee 
and your staffs to advance the effort to modernize the Internal Revenue 
Service.
    I am Paul Cofoni, President Federal Sector, and Corporate Vice 
President for Computer Sciences Corporation(CSC) headquartered in El 
Segundo, California. With approximately92,000 employees worldwide, CSC 
is a world leader in helping our clients, both government and large 
business, use information technologies to achieve business objectives. 
These services include systems integration, consulting, and change 
management.
    I am here today to provide you a status of the Business Systems 
Modernization Program at the Internal Revenue Service. Since December 
of 1998, we have led the PRIME Alliance for the Internal Revenue 
Service. The PRIME Alliance includes some of the best names in the 
technology and business modernization world: SAIC, IBM, Unisys, 
Northrop-Grumman and BearingPoint. While there have been some 
difficulties with Business Systems Modernization or BSM as it is 
referred to, the PRIME Alliance, in partnership with the IRS, have 
created a strong program, architecture, and technology foundation for 
future success. But, before I talk about both the successes and 
difficulties at the program, I would like to touch briefly on the 
history of how we got to where we are today.
    Past Attempts at Modernization--The IRS originally developed its 
database systems in the late 1950s and 1960s to capture, store, and 
process tax return and payment information. These systems, known as the 
``Master Files'', were developed largely on mainframe platforms that 
provided the requisite performance, capacity, and security. At the 
time, the IRS computer system was widely viewed as leading the world in 
the automation of tax collection. As the volume of data mushroomed over 
the succeeding decades and as federal statutes evolved concerning the 
privacy and nondisclosure of confidential taxpayer information, the IRS 
experienced greater difficulty in managing the data. The difficulty has 
intensified with the increased demand for online data to resolve 
taxpayer account issues, facilitate examination and collection, as well 
as provide for improved taxpayer service.
    Two attempts to modernize the IRS technology base failed. During 
the 1970s, the IRS, limited by funding constraints, was only able to 
replace worn out computers. Because of the underlying architecture was 
so antiquated, it limited the IRS's ability to access taxpayer account 
information. To overcome this shortcoming and to augment the delivery 
of taxpayer services and compliance functionality, the IRS developed 
stand-alone ``stovepipe'' systems with separate databases. While these 
systems provided some access to limited taxpayer account information, 
as the system evolved and more demands were made, the IRS experienced 
increasing difficulties synchronizing disparate stand-alone systems and 
databases. The IRS computing environment evolved into an 
extraordinarily complex array of legacy and stand-alone systems with 
respect to both connectivity and interoperability between the mainframe 
platforms and the plethora of distributed systems. Additionally, 
maintenance and annual updating of tax changes became increasingly more 
expensive and risky.
    In 1983, Congress approved a comprehensive technology improvement 
plan, called Tax Systems Redesign (TSR). IRS haste to introduce new 
technology for the 1985 tax season was in large measure responsible for 
the first-ever filing season failure at a cost of $15.5M in interest on 
delayed refunds. In response, Congress approved a sweeping Tax System 
Modernization (TSM) program that was projected to cost $4B and was 
slated to be operational by 2000. The Treasury Department dismantled 
TSM in 1996 after repeated reviews by GAO and this committee pointed 
out that the program was not delivering any significant business and 
processing improvements.
    The PRIME Contract--The 1998 Restructuring and Reform Act mandated 
that the IRS focus on serving the public and meeting taxpayer needs and 
paved the way for Business Systems Modernization (BSM). BSM is one of 
the largest civilian technology renovation programs ever to be 
undertaken. This modernization effort involves massive, long-term 
change for all IRS organizations and for taxpayers.
    Shortly after passage of the Restructuring Act and after almost two 
years of competition, the IRS awarded the CSC PRIME Alliance team the 
Prime Systems Integration Services Contract (IRS PRIME) in December 
1998. It is a 15-year, $5B to $7B, Indefinite Deliver, Indefinite 
Quantity (IDIQ) contract, where work to be performed is identified by 
task order and separately procured. Through December 2003, Congress has 
appropriated approximately $1.35B for the BSM program. Of the $1.35B, 
the IRS has funded CSC for approximately $927M through 114 task orders 
and 1100 task order modifications.
    CSC, in its role as the prime integrator (The PRIME) for 
Modernization:

      Provides program management, technical, and process 
infrastructure necessary to acquire and integrate business solutions 
into the evolving IRS operational environment;
      Assumes lead responsibility for maintaining the 
architecture and standards for Modernization, validating business 
requirements, reengineering business processes, preparing business 
cases, and developing alternative engineering solutions;
      Acquires, integrates, tests and deploys modernized 
systems together with organizational change activities, business user 
training and other support functions;
      Competitively selects best-value technology solutions 
that are derived from commercial best practices and custom-off-the 
shelf (COTS) products developed by our PRIME Alliance partners and 
other contractors;
      Uses its established commercial methodologies and best 
practices (CatalystSM) and those of our partners to manage 
the contract and modernization efforts; and
      Also may provide post-production systems operation and 
maintenance (O&M) support to enable the IRS to leverage PRIME's 
expertise and procurement flexibility with the aim of enhancing the 
post-production Modernized environment.

    Business Systems Modernization Account--To manage funding for 
modernization, Congress established the Business System Modernization 
Account and created several strong pre-conditions for the release of 
funds from the account. IRS was and continues to be required to 
undertake the following:

      Create and continuously implement an Enterprise 
Architecture (EA) that is an institutional blueprint defining how the 
IRS operates today, in both business and technology terms, and how it 
wants to operate in the future;
      Develop and follow a lifecycle management program;
      Acquire the services of a prime contractor to lead the 
system integration effort; and
      Submit to frequent, in-depth audits and reviews by the 
GAO, the Treasury Inspector General for Tax Administration (TIGTA), and 
the Office of Management and Budget (OMB).

    As a consequence, unlike past efforts, every dollar expended under 
the BSM program must undergo one of the toughest oversight and 
compliance process in the Federal Government. These include stringent 
business case development, compliance with the program architecture, 
and the meeting of specific and defined milestones for each module of 
the overall program, before additional funds can be released from the 
account. While at times a burdensome and time-consuming process, it is 
the government's insurance policy that its investment will be spent 
wisely on technology solutions that deliver value to both the 
government and the American taxpayer.
    Significant Accomplishments--As stated earlier, the PRIME Alliance 
and the IRS have delivered key program and technology foundational 
elements for the BSM program. Additionally, a number of business 
applications have provided significant business value to IRS employees 
and more importantly, the American taxpayer. I want to take a moment to 
summarize the joint accomplishments of the PRIME Alliance and the IRS 
since the inception of the program. Let me begin with the program 
foundational successes.
    Enterprise Architecture/Transition Strategy. Building on the IRS-
developed Blueprint for Technology Modernization, published in May 
1997, CSC and the IRS jointly developed the Enterprise Architecture 
(EA)in January 2001. The EA is a business and technology blueprint that 
defines both the IRS future state and the approach to achieving it. 
This joint effort by CSC and the IRS has resulted in recent 
recognition. The EA received the Excellence in Architecture Award from 
the E-Gov, FCW Media Group and Federal Enterprise Architecture 
Certification Institute on 14 September 2003. Its companion document, 
the Enterprise Transition Strategy (ETS) provides the roadmap and 
schedule for implementing the components defined in the EA. The ETS is 
updated yearly in response to changing priorities and budgetary 
constraints, and provides CSC and the IRS with critical information for 
use in making business investment decisions.
    Enterprise Lifecycle Process. A key foundation element for program 
success is the creation of a methodology, tailored to the environment 
and needs of the IRS, for implementing the EA. Using our own 
proprietary methodology, CatalystSM, as the foundation, we 
worked with the IRS to create the Enterprise Lifecycle (ELC), an 
approach that integrates business and technical change in the IRS and 
is responsive to the IRS effort to simultaneously change its business 
enterprise and its IT systems. The ELC is mandated for all 
modernization projects.
    Software Acquisition Capability Maturity Model. CSC adopted the 
Software Engineering Institute's (SEI)Software Acquisition Capability 
Maturity Model (SA CMM) as a program management model. Following 
deployment and internal benchmarking, CSC became the first organization 
in the world to be evaluated at Level 3 of the SA CMM. Leveraging this 
success, CSC assisted the IRS in preparing for its SA CMM evaluation. 
In December 2002, the IRS was evaluated at a Level 2, the first Federal 
civilian agency to achieve this level. All PRIME Alliance partners are 
required to be rated at Level 3 or higher of the SEI Software 
Development (SW) CMM.
    Security and Technology Infrastructure. In May 2002, CSC deployed 
the Security and Technology Infrastructure Release (STIR) that provides 
a common, modernized IT infrastructure for secure interaction between 
employees, tax practitioners, and taxpayers. With the need for an 
extraordinarily high level of security to protect the integrity of 
financial and taxpayer information, deploying modernized applications 
today and in the future is not possible without the STIR in place. This 
hugely complex firewall will enable the IRS to fulfill the 
congressional vision of ``a customer focused IRS,'' that can provide 
taxpayers with many self-help, Internet based options for dealing with 
a complicated tax system, while at the same time providing the highest 
level of security of confidential taxpayer information.
    Enterprise Systems Management. Additionally, the PRIME has 
delivered a modern tool to assist the IRS in managing the health and 
security of the entire technology system. Enterprise Systems Management 
(ESM)capabilities provide an around-the-clock systems monitoring to the 
e-Business Modernization applications. The ESM foundation was laid for 
a centralized enterprise-wide management system that will identify in 
real time lapses in systems performance enabling the IRS to act 
immediately to provide high-availability of critical IRS taxpayer 
applications.
    Let me now turn to the business solutions that have been 
implemented over the last three years that have provided real business 
value to the IRS and to taxpayers and third parties who interact with 
the IRS by telephone or through the Internet. These solutions not only 
provide improved services levels to taxpayers as desired by Congress 
but the solutions in place today can be leveraged in the future to 
continuously provide improvements to service levels faster and less 
costly. A clear example of how one application was leveraged to 
implement a program quickly and less costly was the Advanced Child Tax 
Credit legislation passed by Congress last summer. I will touch on this 
successful implementation later in my remarks.
    Customer Communications 2001. This project improved the IRS' 
telephony architecture by implementing intelligent call routing 
technologies. CC01 increased the number of taxpayers serviced through 
efficient call routing and shorter wait times to reach the appropriate 
customer service representatives (CSRs). CC01 also implemented voice 
recognition for English and Spanish callers and delivered telephony 
improvements that nearly double the capacity at the 25 IRS Call Centers 
from 800 to 1500 calls per hour. This improved capacity allowed the IRS 
to handle 46 million calls in four months during the 2003 tax-filing 
season. Approximately 84 million calls were routed in fiscal year 2003. 
Today, the IRS is experiencing a 50 percent reduction in abandoned 
calls and wait time and the number of Spanish calls has doubled.
    Internet Refund Fact of Filing. In 2002, we expanded the IRS's 
customer communications capability when we deployed Internet Refund 
Fact of Filing, (IRFOF), a Web-based application that allows all 
taxpayers online access to account information and the ability to track 
the progress of their tax returns, including refunds. Performance of 
IRFOF has far surpassed original expectations. Two million hits per 
year were projected; IRFOF handled more than 15 million requests in its 
first tax-filing season, and 17.9 million requests in 2003. 32% of all 
refund inquires came through the IRFOF Web page. As I stated earlier as 
an example of how current applications can be leveraged, in July 2003, 
we leveraged the IRFOF application by implementing the Advance Child 
Tax Credit (ACTC) application. Built in 2.5 months, ACTC provided 
online access to tax creditpayment status to about 26 million taxpayers 
and about 15.5 million inquiries have been received to date. This is a 
success story that received media attention on August 8, 2003 in 
Government Computer News. The Headline read . . . Taxpayers rushing to 
IRS.gov . . . Let me quote from the article because I really believe 
the article best describest the business value of the solutions that 
CSC is delivering to the IRS and taxpayers.

         One of the most popular Web sites last week was not a sports 
        or entertainment site, but--www.irs.gov. Eager taxpayers trying 
        to determine how much and when they would receive their child 
        tax credit checks swamped the site. The spike in traffic made 
        the IRS one of the top 10 Internet sites for the week ending 
        July 27, said Max Heineman, a spokesman for Internet traffic 
        researcher Nielsen/NetRatings.
         . . . About 9 million visitors used the IRS' ``Where's my 
        advance child tax credit'' feature between July 14, when it 
        went live, through yesterday, said IRS spokesman Tim Harms. 
        ``It's as successful as the `Where's my refund' feature was 
        last filing season,'' he said

    Customer Relationship Management Exam. This project modernized 
policies, processes, and technology to enable faster case resolution 
and higher customer satisfaction. CRM Exam deployed an off-the-shelf 
case management and resolution tool, Bureau of National Affairs (BNA) 
Corporate Tax Audit Analyzer (CTAA). CRM Exam allows for highly complex 
tax computation automatically, thus increasing confidence in revenue 
agents' data, while reducing exam time. The average time spent on tax 
computation was reduced from 53 to 17 hours, a 68% reduction. The 
project team trained nearly 4,000 agents in the Large and Mid-Size 
Business (LMSB) operation in use of the application with a training 
approval rating from agents of 82%.
    Internet Employer Identification Numbers (I-EIN). Deployed in April 
2003,this project enables employers and tax practitioners to apply for 
and receive employer identification number online quickly and securely, 
and with less direct involvement by the IRS. I-EIN decreases taxpayer 
burden through 24-hour availability and elimination of paper forms. 
About 453,000 EIN applications have been processed since this 
functionality was implemented.
    e-services. Delivered over the last 7 months, the e-services 
project offers a suite of Internet-based applications providing 7 
distinct business capabilities to electronic return originators and 
third party practitioners. These capabilities will answer the needs of 
Electronic Return Originators (EROs) and third parties who interact 
with the IRS almost daily and who have demanded these real time 
paperless services for some time. Here are the new services that are 
now available. EROs can now register electronically to do business with 
the IRS as an ERO; they can now submit Power of Attorney applications 
electronically reducing the time to represent their clients before the 
IRS. The Transcript Delivery Service will provide real time electronic 
delivery of tax return and account information to tax practitioners and 
other third party users such as State and Federal agencies. e-services 
solutions also address recurring operational issues within the IRS, 
including lengthy cycle times and high percentage of rework, by making 
it easier for taxpayers and other entities to transact business with 
the IRS and by providing faster responses. Here is an example of how 
the new e-services functionality will streamline one area of operation 
of the IRS. Currently the IRS receives 1.4 Billion transactions each 
year with individual names linked to taxpayer identification numbers. 
Approximately 70 Million have errors. e-services will allow submitters 
to validate the tax ID number before submission, dramatically improving 
data quality.
    Near Term Deliverables--During 2004, the PRIME Alliance and the IRS 
are going to deliver a number of important projects providing 
significant benefits to taxpayers and to the IRS in improving overall 
management of the financial area of the agency.
    Integrated Financial Systems. When deployed, this project will be a 
key enabler to ensure that the IRS meets all internal and external 
requirements for management controls, performance measures, and 
financial reporting. The IFS will correct material deficiencies in 
current financial processes identified by GAO and will help the IRS to 
sustain an unqualified audit opinion on its consolidated financial 
statements and comply with legislative directives. The IFS provides for 
a single, integrated source for budget management, core accounting, and 
cost management data and provides a general ledger for custodial and 
administrative accounting. The IFS is based on an industry-standard 
COTS financial package (SAP) with tailoring (configuring of the 
package) to address unique Chief Financial Office (CFO) requirements. 
The initial release, including core financial and budgeting 
functionality is scheduled for delivery before October 2004.
    Modernized e-file. Within the next few months, the IRS will be able 
to receive corporate tax returns and information returns 
electronically. This important step will help move the IRS closer to 
the 80 percent goal for electronically filed returns while promoting 
error-free filing and immediate access to return information for use by 
taxpayers and the IRS. While CSC was not the integrator for this 
project, this solution will run on the infrastructure platform built by 
CSC and the functionality from one of the e-services applications 
delivered by CSC will became a key component of this project.
    Customer Accounts Data Engine--The Customer Account Data Engine 
(CADE) is regarded as the most critical building blocks in the entire 
BSM program. CADE will replace the 35 year-old master file system that 
contains the authoritative record of all taxpayer accounts. The current 
system is extremely large and in constant use. It requires 
approximately 13 terabytes of mass storage and during a peak week 
performs 28 million transactions. CADE will replace the existing system 
with new technology, new applications, and new databases to provide IRS 
employees with real-time, electronic access to all aspects of a 
taxpayer record within mandated requirements for the security and 
privacy of taxpayer data. Once fully operational, taxpayers will see 
real benefits as well. Today's system is designed to process return 
data in a weekly cycle; the new system will reduce the cycle time from 
one week to one day, thus allowing faster refunds to taxpayers and also 
making taxpayer return information available sooner. This will be 
extremely important as the IRS moves closer to a self service strategy 
for taxpayers, tax practitioners and other third parties.
    Where are we today with CADE implementation? That first release of 
CADE was completed in early January 2004, which I consider a major 
milestone for this program. This milestone represents a lot of hard 
work and dedication by the joint IRS/CSC PRIME team to make this event 
a reality.
    Many of you, I know, have read press accounts around the delays in 
delivering CADE. As I have said before, CSC takes full responsibility 
for its share of the delay in completing the first release of CADE. We 
at CSC did not understand the complexity of the current systems 
environment; nor did we understand what it would take to build a new 
data base platform for the IRS. But let me say now, however, that in my 
30 years of working in the technology field, I have never encountered 
any program of the size and complexity as the business systems 
modernization program at the IRS.
    To put this program in its proper perspective: a 60's system, 
largest data base of its kind (150 million taxpayers), $2 trillion 
payments processed annually, about 250,000 business rules (about 50,000 
business rules around the first release of CADE), intense security 
requirements, 60 million phone calls received annually during the 
filing season, and added complexity of writing software that will allow 
the IRS to operate both systems simultaneously during the transition 
period.
    Can you think of any organization in the world that would match the 
size and complexity of the IRS? I cannot.
    With the first release of CADE complete, we are now on a path to 
complete the next release by mid 2004 and during the 2005 filing 
season, the Customer Account Data Engine will process the first 1040EZ 
returns filed by over 6 to 7 million taxpayers.
Schedule and Budget Problems
    While the program has experienced delivery issues with schedule and 
cost, I believe that the program is on a path that will demonstrate 
substantial improvement in delivery performance in the near future. Let 
me tell you why I believe this to be true. The top leadership at CSC 
and the IRS are committed and focused on building a business systems 
program second to none. Over the last several months, CSC and IRS 
senior executive leadership now meet regularly to discuss program 
performance. Mike Laphen, our COO and Commissioner Everson meet monthly 
and those meetings have resulted in discussions around the issues that 
impede progress on the program and reaching agreement on the steps 
needed to continue the momentum that the program has experienced. And 
every two weeks I, along with the CSC General Manager of the Program, 
Jim Sheaffer, meet with John Dalrymple the Deputy Commissioner for 
Operations Support, Todd Grams, the CIO and Fred Forman, Associate 
Commissioner, Business Systems Modernization, to discuss project status 
along with other critical issues around the program. Additionally, we 
have begun a process of co-locating CSC and IRS executives to improve 
communications and to create a closer working relationship, 
characteristics so important to successful implementation of a program 
of this size and complexity.
    Since July 2003 a number of external studies of the program have 
been conducted at the request of Commissioner Everson and our COO, Mike 
Laphen. Soon after his appointment, Commissioner Everson engaged 
Carnegie Mellon University, Software Engineering Institute (SEI) to 
perform a ``health check'' on the proposed use of the Sapiens eMerge 
business rules engine for the CADE project as well as to conduct a full 
Independent Technical Assessment of the CADE program. Such assessments 
are conducted in response to cost, schedule and performance problems.
    Principal findings of the SEI report concluded that the PRIME now 
has the technical and management talent to deliver the first release of 
CADE. That finding has become a reality as the first release was 
completed in early January, as I stated in my earlier remarks around 
the CADE project. Another principal finding was that the business rules 
approach executed by CSC PRIME is conceptually sound but that the 
technology needs further evaluation. The evaluation process for 
business rules has already begun.
    To put the accounts of schedule and cost slippages into 
perspective, I want to take a moment to make you aware of a key point 
made in the SEI report around the complexity of the first release of 
CADE. The report observed that CADE was ``in uncharted waters'' for the 
IRS/PRIME team. Moreover, the report stated that:

         ``Early in the assessment, both the PRIME and the IRS asserted 
        that Release 1 was the ``simplest of releases--just the 1040EZ 
        . . . the `it's the easy release' public face undermined the 
        ability of stakeholders, including Congress, to grasp the 
        complexity of the release, where approximately 85% of the code 
        is CADE infrastructure, with the remaining 15% related to the 
        1040EZ business rules. A better approach would have been to 
        publicize Release 1 as the foundation for CADE to better set 
        stakeholder expectations.''

    I think this independent observation by SEI is an important lesson 
for both CSC and the IRS. The lesson for me is that all of us connected 
with the program must convey the``right'' message to our stakeholders 
about the business of modernizing the IRS. We obviously have not done 
that in the case of the CADE project.
    At the same time as the SEI assessment was underway, my COO made 
the decision to invest in an outside study to evaluate the program with 
a focus on the significant inhibitors to on-time and on-budget 
performances, and to make recommendations on what we can do to improve 
overall delivery performance. Bain and Company was selected to perform 
this analysis.
    Bain and Company identified two primary drivers for the problems 
around on-time and on-budget performance: loosely defined requirements 
and under performing execution of the projects. Their review concluded 
that significant changes are needed to successfully deliver business 
systems modernization to the IRS. First, we must improve significantly 
the business requirements definition process. We can no longer wait 
until a project is in the testing phases of a project to identify a new 
business requirement. Second, both CSC and the IRS must streamline the 
decision-making process and authority. The program needs a single 
source for decision-making; concenus decision-making among many is not 
feasible for a program of this complexity. Third, both parties must 
have an increased focus on business transformation, including more 
business representation on project teams, and an increased role by the 
business in requirements definition and transition planning. Lastly, 
improvement in subcontractor accountability and delivery is essential 
for success. We must hold our subcontractors accountable for excellent 
performance.
    In late October 2003, Commissioner Everson and COO Mike Laphen 
initiated 46 action plans to address the findings of these studies over 
a period of six months. Senior executives from the IRS and CSC are 
leading the work around these action plans and as of today, work has 
been completed on almost two-thirds of the actions. The IRS Oversight 
Board issued an independent report in late December and the principal 
recommendations in that report are covered by the joint IRS/CSC action 
plan work.

Closing Remarks
    In closing, the IRS Modernization Program is at the top of my 
company's watch list of projects. Our CEO is briefed periodically on 
the progress we make in modernizing the IRS as well as the challenges 
we face in delivery performance. COO Mike Laphen, Jim Sheaffer, the 
General Manager, and myself are personally committed to program success 
and we will do what it takes to deliver. One study, SEI, stated, ``stay 
the course.'' I cannot agree more. We have made significant progress 
over the last 6 months. The fact is we have delivered significant 
business value to our American taxpayers and to the IRS. The 
infrastructure that we have built is stable and secure and will serve 
as a cornerstone for future successful deliveries of improved services. 
We have created a solid management process foundation, and we have 
delivered a number of applications that ease the burden of taxpayers 
and third parties who interact with the IRS by telephone and through 
the Internet. While I know that more challenges lie ahead, I also know 
that the CSC team in place can successfully meet those challenges. The 
IRS and CSC must now leverage our past experiences with modernization 
and continue the commitment to work together in partnership and trust 
(and I cannot emphasize enough the importance of partnership and trust) 
to improve overall delivery performance and to eliminate on-time and 
on-budget issues. The goal of CSC, as the integrator for this important 
program, is to deliver the best tax administration system to the 
American taxpayers. I am committed to achieve this goal in partnership 
with the IRS and Congress and our various external stakeholders.
    Mr. Chairman, I thank you for the opportunity to appear today 
before you and your subcommittee and I will be happy to answer any 
questions.

                                 

    Chairman HOUGHTON. Thank you very much, Mr. Cofoni. I am 
just going to ask a question, and then I will turn it over to 
you, Earl and Rob.
    I thank you very much for your testimony. Let me just 
direct this to you, Mr. Cofoni. You have got a great company, 
and you have done a lot of things right. At the same time, we 
are responsible for people and their money and the investment 
of the tax dollars. The bottom line is that we haven't met the 
mark. We can't manage this. We are the distributors of funds to 
the IRS, but we have got to have oversight and confidence that 
this is going well.
    It seems to me that there is such a difference between your 
testimony and that of some of the others. Tell me, why is this 
difference there?
    Mr. COFONI. I don't think there is a fundamental 
difference. I think we are proud of the accomplishments that 
our people have made and our alliance partners have made, and I 
wanted to make sure that we articulated those for you today to 
make sure you had a balanced view. We accept responsibility and 
accountability for the slippages in the case of both CADE and 
now IFS. By the way, of all the five projects that are being 
discussed, three of those fall under my contract, our contract.
    For the two, CADE and IFS, we have stepped up and we have 
taken the responsibility for cost overruns. We did that for 
CADE a year and a half ago and we did that just recently when 
we announced the IFS overrun. So, we recognize that these are 
problems, we have taken aggressive action based on the studies 
that have been done, and we will continue to take actions to 
improve.
    There is no escaping the fact that we are not proud of our 
performance on the program in terms of cost and schedule. We 
are quite proud of the deliveries, the quality of those 
deliveries, the acceptance within the IRS and by taxpayers of 
those deliveries.
    Chairman HOUGHTON. I think when a contractor takes on a 
job, it writes its name in blood on this thing.
    Mr. COFONI. That is true.
    Chairman HOUGHTON. This is what we are going to do, and if 
we don't do it, it is our fault. If we don't do it, you have 
got to have an interaction between the people who you are 
working for and you have got to anticipate and make tough 
demands. If this isn't working, we are going to do this, and if 
that isn't right, we can't do the work at all.
    I don't see that. I see it has been bubbling and bubbling 
and bubbling along, and all of a sudden, there is 
disenchantment within the system and it is laid on your 
doorstep and really, in effect, you are the fellow that is 
responsible.
    Mr. COFONI. Yes, sir. I accept that criticism, and we have 
put some new stakes in the ground. We are not going to start 
new work before the requirements are fully defined in detail. 
That has been at the heart of our problems. The core problems 
underlying these overruns have to do with unknown requirements 
and unknown data conditions. There are other contributors, as 
well. I don't mean to single only those, but those are the 
core.
    We have rededicated ourselves to not go forward. We will 
not take work if we don't have requirements defined well in 
advance. We will walk away from that work.
    Chairman HOUGHTON. That is ex post facto. What about a year 
ago?
    Mr. COFONI. I think a year ago, our people were trying to 
do what they believe was in the best interests of the IRS. If 
we didn't feel we understood all of the requirements, we moved 
forward in any--we moved forward. We moved forward with the 
feeling that we were doing the right thing for the IRS. As it 
has turned out, this was not the right thing. The right thing 
would have been to have stopped work at that point, not be so 
accommodating, and demand that we had a detailed sense of 
requirements, that we put our absolute best people into the 
requirements process collectively and drove those out.
    I will add, however, even after doing that, these are 40-
year-old systems. When I ask my people about the documentation, 
they laugh. The documentation is not there on these systems. 
So, in many cases, our requirements, defining the requirements 
that are embedded in the existing systems is like an 
archaeological dig. We have to strengthen our interrogation 
processes and our research processes to get at a great 
percentage. I think we will always need a reserve against those 
unknown conditions that no one seems to have----
    Chairman HOUGHTON. Rather than laughing, did you point this 
thing out to the people in the system?
    Mr. COFONI. Oh, of course. They take this seriously. This 
reflects on their personal performance, their careers. They are 
in most cases working 6 and 7 days a week, 60 hours or more, 
100 percent committed to the effort, but we have them in a very 
difficult----
    Chairman HOUGHTON. Well, look, we will have a chance to 
kick this around and get the opinions of other people on the 
panel. I would like to ask Mr. Pomeroy to take over here.
    Mr. POMEROY. Mr. Dacey, I think I will ask my first 
question to you. In looking at a table of cost overruns, it 
appears that the magnitude of what we are talking about is cost 
overruns of $290 million, a cumulative delay in terms of 
deadlines blown of 83.5 months. Now, that is taking specific 
programs and adding them together, but does that sound about 
right in terms of what you have been able to see with this 
project?
    Mr. DACEY. Yes. The table in our testimony is two parts, 
first of all, the completed projects, which we talked about 
earlier today, as well as the ongoing projects. It doesn't 
count some of the other efforts that were taking place in prior 
years for which those projects have been deferred or delayed 
for the future.
    So, with respect to our table, this does represent the 
overruns for those projects, but again, there are other 
projects that aren't on this table that have been part of the 
earlier parts of BSM.
    Mr. POMEROY. So, as bad as this is, if you go back in time 
just a bit, it gets worse?
    Mr. DACEY. Well, there are other costs in there and we 
haven't analyzed the overruns, but in going through this 
process, we had made recommendations consistently that they 
needed to balance the pace of these projects with their 
capability to manage them. As part----
    Mr. POMEROY. Thank you. I am sorry, I didn't mean to cut 
you off, I just have--I want to use my time as well as I can 
here.
    Mr. Levitan, you have brought a career's worth of 
experience in consulting, looking at relationships between 
enterprises and their consultants, looking at consultant 
contributions to major project upgrades within enterprises. 
Based upon the wealth of experience you bring to the Oversight 
Board, how would you describe this particular project, as way 
off, really horrible, missed the mark a bit? Where in the 
spectrum are we?
    Mr. LEVITAN. First of all, it is a very, very difficult 
program, but that doesn't make excuses. We have missed the mark 
significantly. The IRS has missed the mark in managing the 
program. The PRIME contractors have missed the mark both in 
delivering results based on commitments for target dates and 
costs and also in their responsibilities to be a trusted 
partner and advisor to the IRS and help the IRS in an effective 
way in managing the programs. It has been a significant miss on 
all accounts.
    Mr. POMEROY. I appreciate that comment. I do think it is 
important that we underscore the scale of what we are talking 
about. Some of the discussion sounds like an unsatisfactory 
performance review that might be conducted within the norms of 
business operations, but I believe missing the mark by $290 
million, the delays, the insufficiency of meeting the deadlines 
is really of a stunning magnitude. If this does not provide 
some breach in the trusted relationship between contractor and 
the IRS, I wouldn't know what was. Certainly looking at it from 
the generalist perspective of a Subcommittee on Oversight 
Member, I am stunned by what I am seeing here, and I am deeply 
alarmed about it.
    Mr. Palmquist, some of your evaluation is that the 
technical expertise with the contractor wasn't quite where it 
needed to be to get this job done, if I understand your 
testimony correctly.
    Mr. PALMQUIST. Yes. We had questions about, say, things 
like the testing process which the Department of Treasury 
Inspector General also had questions with. Frequently, testing 
was not coordinated, say, between CSC and IBM, where they may 
both be looking at the same defect, both approaching a 
solution, but not in a coordinated fashion. So, they may both, 
in fact, be correcting the same thing, and then those 
corrections may not sync up later on. So, in many cases, it was 
not coordinated on a technical side.
    Mr. POMEROY. If I understand correctly, in order to make 
certain that you have got sufficient horsepower in your 
contract, in this case, there is a benchmark that they need to 
make, a capability maturity model (CMM) certification. Is that 
correct? Is that what this is geared toward?
    Mr. PALMQUIST. We have several CMMs, the software 
acquisition CMM that Mr. Cofoni spoke about, as well as the 
software development CMM. The CMM is only part of the solution, 
to take the SEI itself as an example. The SEI has four 
initiatives--process is one of them. The other ones are 
architecture, security, and performance critical systems. 
Processes in and of themselves don't result in a defined 
product, as I stated in my testimony. Sometimes these processes 
were not backed up by sufficient technical experience or 
expertise, and that was one of the areas where we found some 
lacking.
    Mr. POMEROY. So, is CMM a process evaluation or does it 
also include personnel and their competence?
    Mr. PALMQUIST. It is an evaluation of processes.
    Mr. POMEROY. Although Mr. Cofoni notes that they are the 
first to obtain the Level 3 certification for processes under 
the CMM, I would note that you just got it last August when 
under the contract you were supposed to have it July 1, 1999, 4 
years earlier. To this point, Mr. Cofoni, did you receive any 
financial penalty for being 4 years late in having your system 
certified as required under the contract?
    Mr. COFONI. No, sir.
    Mr. POMEROY. I am interested in what is the manner of 
financial penalties you have received for failing to meet 
performance, for contractual commitments.
    Mr. COFONI. We have, since 18 months ago, some 18 months 
ago, been paying for all of the work being done on CADE. All of 
our costs and our PRIME Alliance partners' costs have been paid 
for by our company.
    Mr. POMEROY. What have you been compensated in cost 
overruns under your contractual relationship with IRS, do you 
know?
    Mr. COFONI. I don't have that. About four----
    Mr. POMEROY. It was cost overruns of $290 million, and you 
having the, certainly the lion's share of the relationship, I 
would expect most of the compensation under the cost overruns 
has come to your firm, correct?
    Mr. COFONI. That is correct for those projects on the list 
that are within the scope of the PRIME contract, which are the 
ones--I don't have the list in front of me--so the ones that 
are active projects today are e-services, which we just 
concluded, IFS, and CADE. The others, I don't have the table 
that you are referring to so I am a little disadvantaged, but--
--
    Mr. POMEROY. While your people review that, I have got 
another question that actually you may find more agreeable. You 
indicate that under this contract, you have discharged 114 task 
orders and 1,100 task order modifications. Now, are those 
essentially change orders?
    Mr. COFONI. Yes.
    Mr. POMEROY. That you are getting from IRS?
    Mr. COFONI. Yes.
    Mr. POMEROY. The IRS is not represented here, but someone 
looking over this, either Mr. Dacey or any of the other three 
of you, are these change orders driven by legislative changes 
that Congress keeps passing so that the target keeps moving of 
what we are trying to get the system to do, or is it simply a 
very, very poorly commenced project?
    Mr. LEVITAN. Mr. Pomeroy, let me respond to that. The 
change orders are many different things. Some of them are 
legislative in nature. Others are that the IRS did a poor job 
in defining their requirements in the beginning and then their 
processes of controlling the change orders was not very 
effective. They are trying to address that now by the 
organizational changes that the Commissioner described 
previously, putting Mr. Dalrymple in charge of managing that 
process.
    Again, many of those change orders, I would say the 
majority of them were initiated by the IRS. Some of them were 
absolutely necessary. Some should have been caught much 
earlier. I would say that the issue there rests primarily with 
the IRS, not with the contractor.
    Mr. POMEROY. Is there a broader lesson to be drawn from 
this? As we outsource, we absolutely must retain within the 
staff structure of the government agency doing the outsourcing 
sufficient technical competence to adequately engage and 
oversee the contractor?
    Mr. LEVITAN. You are absolutely correct. It is not just 
technical competence. Even more importantly than that, it is 
the project management competence. Quite honestly, the Board 
has been telling the IRS that its own capabilities to manage 
the program and oversee the contractors has been inadequate. 
They have been very slow to move on that and make the necessary 
improvements. Again, steps are underway at the present time to 
bring in additional resources to help accomplish that.
    Mr. POMEROY. A final focused question, Mr. Chairman. Thank 
you for your leave here. Mr. Cofoni, you indicate in your 
testimony you have 92,000 employees worldwide. Is this work 
being done in-country?
    Mr. COFONI. All of the work for the IRS is being done in-
country, principally done in Maryland in our New Carrollton 
facility across the street from the IRS facility.
    Mr. POMEROY. Thank you.
    Chairman HOUGHTON. Thank you. Mr. Portman?
    Mr. PORTMAN. Thank you, Mr. Chairman, and thank you, 
gentlemen, for your testimony this morning. I started off my 
last comments saying here we are again, and it really is 
frustrating for all of us to once again be in a situation where 
we are partway through a business modernization process, in the 
case of the important work on CADE and important work on the 
IFS. We were hoping to get IFS done early this spring, hoping 
to get CADE done back in year-end 2002, I believe, and yet we 
are reaching to try to get those done this year. So, the 
question is, how do we keep the momentum going toward a 
deliverable here, at the same time being sure that the problems 
we have encountered are handled properly.
    The first question I am going to ask is to the entire panel 
except for Mr. Cofoni, which is should we change PRIMEs at this 
point? Are we at the point where, again, despite the fact that 
at least in a couple of these deliverables we are close, we 
hope, to accomplishment, have we had enough problems here in 
terms of the cost overruns, in terms of the delays, that we 
should change the PRIME? Mr. Levitan, I will put you on the 
spot.
    Mr. LEVITAN. Mr. Portman, I am going to have to come back 
at you. The RRA 1998 specifically precluded the Board from 
getting involved in procurement activities. Our legal counsel 
has told us that advising the IRS to fire a contractor would be 
getting ourselves involved in that.
    We have been very strong in saying that the IRS needs to 
look at all options in strengthening the team to get done what 
needs to get done, and we feel very strongly about that. We are 
precluded from making a recommendation to fire anybody.
    Mr. PORTMAN. I have to respect your inability to get 
involved in procurement, since those of us on this panel were 
part of establishing that and I think that was appropriate. I 
am not sure that it is fair to say that we didn't view the 
Board's authority to include looking at a big picture issue 
like this and telling us what direction we ought to take, so 
perhaps we could differ on whether this is a procurement issue 
or whether it is a recommendation of the Board on a major 
modernization project, but I won't put you on the spot any 
further, at least not in public testimony.
    Mr. LEVITAN. Just let me go one step further. I think that 
looking at this issue is not a black and white issue.
    Mr. PORTMAN. I couldn't agree more.
    Mr. LEVITAN. It is not a ``keep going with everybody doing 
what they have been doing,'' or ``fire the contractors and 
start all over again.''
    Mr. PORTMAN. Yes. There is plenty of blame to go around.
    Mr. LEVITAN. There are a myriad of options.
    Mr. PORTMAN. Part of the blame rests right here, I believe, 
in Congress, because we have not been perhaps as good at 
oversight as we should have been over the last few years in 
following this, being sure the requirements were appropriate, 
being sure there weren't over-promises, being sure that the IRS 
had the management systems. There is certainly blame at the IRS 
and I think you, Mr. Palmquist and Mr. Dacey, have all outlined 
that. I think Mr. Pomeroy just referenced one, which is lack of 
expertise both on management and with regard to technical 
expertise.
    I still pose that question. I am not suggesting it is black 
or white, but I do think that is something that as a fiduciary, 
being a Member of Congress representing a lot of people who pay 
taxes, some of which have now been used for a program that has 
had huge cost overruns, I think it is an appropriate question 
to ask. Mr. Palmquist, could you answer my question?
    Mr. PALMQUIST. Congressman, we, as a federally funded 
research and development center, are also prohibited from 
making direct source selection decisions, but we do, in fact 
provide counsel. We did provide Commissioner Everson our 
thoughts in general on a replacement of a PRIME contractor on a 
contract of this nature and magnitude. The fact is, 
unfortunately in the state of affairs today, many programs, 
with many different PRIMEs, are experiencing similar problems. 
In other words, a new PRIME is not necessarily going to change 
a program radically. Also, a change in PRIME at this juncture 
would result in a tremendous loss of experience that has been 
gained, some good experience, some bad experience.
    So, while we did not and cannot directly advise that, we 
did tell Commissioner Everson that there are a good number of 
issues that would come up in the change of the PRIME. We felt 
it would be a setback in the program of several years.
    Mr. PORTMAN. I won't attempt to paraphrase what you just 
said, but it sounds like what you are saying is you identified 
problems, again, that can be shared, but certainly with the 
contractor, and yet you believe from your experience with other 
agencies and departments that these problems are not unique to 
the IRS. You are not sure there is another PRIME out there that 
has done much better, and that you believe that given their 
experience, it would be a mistake at this point for them to 
pull this contractor.
    Mr. PALMQUIST. Again, sir, without directly making a 
comment that would directly affect the source selection 
decision that this would be, we see a lot of benefit in staying 
the course. We also do see benefit in a change. We see a lot of 
evidence indicating that the team that is in place is a capable 
team, and if corrections are made, can, in fact, deliver for 
the IRS.
    Mr. PORTMAN. Mr. Dacey?
    Mr. DACEY. In terms of----
    Mr. PORTMAN. You are not constrained by any of these 
procurement issues, I know.
    [Laughter.]
    Mr. DACEY. No, but I won't be making a recommendation today 
either on that account. I think the issues raised are valid, 
and certainly SEI in their report raised a number of the issues 
that would have to be considered. Certainly Commissioner 
Everson this morning had indicated an approach to moving some 
of the other contracts away to potentially other contractors 
until it can be demonstrated that CSC can carry out the current 
contracts, and I think those are all valid considerations.
    I think the other issue, too, which hasn't been highlighted 
yet is that a lot of the issues are going to need to be 
resolved by IRS itself and switching contractors isn't going to 
fix that. So, there is a heavy amount of effort that I think 
IRS needs to accomplish, and they have set about doing that 
would have to be done regardless of the PRIME.
    So, I think there are just a lot of issues there. Again, I 
don't have a bottom-line analysis. I won't give one today. We 
haven't studied it in any great detail, but there are 
substantial issues that would need to be addressed and 
considered before any consideration like that were made.
    I would highlight, too, that the Commissioner's 
announcement of looking to other contractors is, from a 
personal standpoint, a little bit of competition, which is 
healthy in that regard. I would, however, warn that it would be 
important to make sure that those efforts, to the extent that 
they interact with the systems that are being developed by CSC, 
are well coordinated and also reiterate our concern which we 
have made for several years that IRS needs to have the internal 
capacities and management capabilities to manage the contracts, 
whether it be by PRIME or someone else.
    So, simply taking that to another contractor may not be the 
full solution. The IRS really needs to make sure they have got 
their house in order and don't take on too many projects that 
exceed their capabilities.
    Mr. PORTMAN. My time is ending and I wish I had much more 
time, but let me just, if I could, ask another general 
question, Mr. Chairman, with your leave. Mr. Cofoni, would you 
like to comment on any of the other three comments?
    Mr. COFONI. No, thank you.
    Mr. PORTMAN. I don't want to put you in that position if 
you are not comfortable.
    Mr. COFONI. We do feel that the right thing for the 
government and the IRS, and this will sound self-serving, but 
we honestly believe that the body of knowledge we have 
accumulated in the last 4 years has enormous value to us going 
forward, so we would obviously--we would like to continue.
    Mr. PORTMAN. It seems to me that, Mr. Levitan, you are 
somewhat optimistic about IRS making some of the changes that 
Mr. Dacey just outlined and that, in fact, you think some of 
those changes have been made even in the short term. Certainly 
the announcement today that the Commissioner is looking to 
other contractors for some of the other projects would be 
consistent with the general advice that the Oversight Board has 
given. You have also, though, given some very specific advice 
on limiting these projects, in fact, even postponing some, 
focusing on, it seems to me, some of the more important ones. 
Do you think the IRS is making progress, and then let me ask a 
general question of the whole panel.
    One of my concerns about this process of contracting on 
information technology is that it seems to me when you go 
through the request for proposal process, which is where these 
companies are competing for this business, that there is often 
an issue with requirements. As we have said, the IRS did not 
perhaps spell out the requirements and some of the data 
surprises may relate to that. Also in the nature of 
competition, there is over-promising.
    I would just like to get, once Mr. Levitan has a chance to 
answer that earlier question, just a general sense, because 
this is important going forward, how much of it is due to the 
requirements not being spelled out properly and true surprises, 
to the extent those can be identified as separate from what 
should have been in the requirements, and how much of it is 
just contractors want this business so they make promises they 
can't keep. Once they get halfway or two-thirds of the way 
through the project, it is their project, understanding that on 
IFS, Mr. Cofoni, you are willing to pick up some of these costs 
yourself. That is just a general question that I have about 
this that I think is relevant going forward. Mr. Levitan?
    Mr. LEVITAN. As I mentioned, the project fell in a ditch 
this summer. I was very impressed with the way the Commissioner 
reacted to that and the way he stepped back and said we have 
got to do a thorough study, we have got to put in a plan of 
improvement.
    Going back to our nine specific recommendations, eight of 
which had nothing really to do with the PRIME, we have seen 
some significant progress on that. Business unit management of 
the program is well underway. Putting John Dalrymple in charge 
of that for the IRS is a very positive step.
    Creating an environment of trust and confidence and 
teamwork is another responsibility of the Deputy Commissioner. 
That is going to take time because it is a cultural issue, and 
even the very--and what I am talking about here is the various 
units of the IRS working together in an effective and trusting 
way. That, I think, is starting to happen. It needs a lot more 
work.
    They have made changes to the systems development life 
cycle to make it more effective and putting in place steps to 
make sure that they are following that, which they didn't 
always do in the past. The contracting process continues to be 
an issue and needs a lot more work.
    The experience of the IRS management team is very 
important. They have started searches for people to bring in 
and strengthen that team. They are doing the right thing, but 
that effort is moving glacially slow and needs to be moved 
ahead much more quickly.
    So, I think a lot of the right steps are underway. They are 
just underway and will require a lot more work to really put 
them in the position that they need to be in to manage the 
program and work with the PRIME in an effective manner.
    Mr. PORTMAN. I know you will continue your oversight to 
make sure that happens, Mr. Levitan. We appreciate what the 
Board does and your expertise.
    On that general question, are there any comments? My time 
is ending here, so you will have to be brief. Mr. Dacey, you 
have got some experience with this.
    Mr. DACEY. In terms of the issues in going forward, again, 
there are a number of challenges. I think that the 
recommendations that were contained in this number of studies 
were good things, and I think IRS's intent to continue to use 
outside folks with expertise to help look at their processes in 
carrying those forward is another important element. These 
studies, which were commissioned in 2003, were fairly extensive 
relative to the work that had been done before from an outside 
viewpoint, and I think continuing that is a very positive 
thing, to keep watching the process as it goes forward and keep 
seeing if it is progressing as IRS plans.
    Mr. PORTMAN. Mr. Palmquist?
    Mr. PALMQUIST. Yes, just to echo, the improvement effort 
itself needs to be treated as a project. It needs to be 
planned. It needs to be budgeted quite honestly and monitored. 
It can't be assumed to happen just by good intent. It needs to 
be its own separate effort.
    Mr. PORTMAN. I thank my colleagues for their deference.
    Chairman HOUGHTON. Thank you. Mr. Weller?
    Mr. WELLER. Thank you, Mr. Chairman. I appreciate this 
opportunity. I direct my question to Mr. Palmquist, Mr. 
Levitan, and Mr. Dacey, if each of you would respond. In its 
report, the SEI identified two cases where the primary 
contractor, Mr. Cofoni's company, attempted to develop software 
in-house instead of acquiring it from companies with expertise. 
These decisions later proved to be costly. I was wondering, 
what would you recommend to avoid a recurrence of this kind of 
decision making?
    Mr. PALMQUIST. Congressman Weller, I assume you are talking 
about TRW and also to the first attempt at a business rules 
engine. The business rules engine effort that failed in early 
2001 turned out to be--we did not investigate this 
extensively--but turned out to be a combination of business 
factors that simply weren't understood at the time and, also I 
think, requirements. The effort to use TRW to do the legacy to 
the computer--excuse me, the interface to the legacy 
environment--we understand that the proposal that was delivered 
did not meet the schedule as envisioned and, therefore, was not 
accepted. In retrospect, that appears to have been a decision 
that cost the program.
    The CSC has worked to invigorate their development 
environment. The use of the business rules engine is part of 
that, but the development environment does need to--is 
improving, but does have some issues still remaining, again, 
with respect to the testing element, and contract and 
requirements management.
    Mr. WELLER. Mr. Dacey or Mr. Levitan, do you have a 
comment, please?
    Mr. DACEY. Again, we haven't done a significant analysis of 
those particular issues that were just discussed, but I would 
like to say that it is important to have appropriate amounts of 
information when these decisions are made, as the TRW example, 
I believe, was a situation where there was, I think, an overly 
optimistic expectation of when this software would be delivered 
and the decision was made, well, TRW can't deliver in that 
timeframe. As it turns out, it has taken much longer to develop 
the product, and as was just mentioned, had that been known in 
the beginning, it could have been factored in most likely.
    So, I think the key gets back to fixing some of these 
fundamental issues of understanding the projects and coming up 
with reasonable cost estimates and processes that will yield 
those. I do think, as said before and we have said before, 
there have been overly optimistic expectations, too, in setting 
these up, as well as the issues having to do with some of the 
other problems that have delayed it, like system requirement 
understanding and things of that nature.
    Mr. WELLER. Thank you. Mr. Levitan?
    Mr. LEVITAN. Systems projects fail or have significant 
overruns for one or a combination of three reasons. Number one 
is we don't have the appropriate methodology. In this 
particular case, we had a pretty good methodology, but it 
needed some improvements. Those improvements have now been 
made, positive factor.
    The second reason projects run into trouble is the project 
teams just don't follow the methodology. They try to take 
shortcuts. That was done over and over again in this particular 
case. We now have a commitment that that will not happen again, 
that we will have the appropriate discipline and management. 
That is yet to be proven, needs to be monitored very carefully.
    The third factor is the skills, capability, and experience 
of the people who are actually doing the work. We think that 
that has been demonstrated, that that has been inadequate and 
has not been fulfilled and that is still to be proven going 
forward, and we are very concerned about that.
    Mr. WELLER. Thank you. Mr. Cofoni, I hope you respect my 
questioning. I am one of those who believes in giving the 
private sector the opportunity to participate and contribute 
because I believe there are efficiencies in the private sector, 
so I hope you realize my questioning was friendly regarding 
that particular issue.
    The other question I would like to direct to you, Mr. 
Cofoni, and you have already addressed part of it, you 
indicated the earlier issue of off-shoring of various types of 
government contracting jobs, that potential, and I know you 
answered earlier that you have no workers outside the United 
States performing any work for this IRS contract. Are you aware 
of any subcontractors or any other contractors to the IRS that 
have that work performed offshore?
    Mr. COFONI. No, sir. All of the work we are doing and our 
PRIME Alliance team members are doing under this contract is 
being performed within the United States.
    Mr. WELLER. Okay. Mr. Levitan and Mr. Dacey, from your 
perspective, there has been some concern that I have heard from 
constituents that I have that there is potential that tax 
preparers may be using workers outside of the United States to 
do tax preparation work and their concern is there is personal 
security, their privacy regarding someone who may have access 
to their personal data when it comes to tax preparation.
    Number one, are you aware of any tax preparation firms that 
today are using workers outside of the United States to do tax 
preparation work, and then from your perspective, are there any 
personal security concerns that we should be aware of or 
sensitive to?
    Mr. LEVITAN. Mr. Weller, the Board has not looked into that 
issue so we really don't know the answer to that. It obviously 
is a concern and could be a concern and deserves looking into, 
but we have no knowledge of this at the present time.
    Mr. WELLER. Okay. Mr. Dacey?
    Mr. DACEY. I would echo Mr. Levitan's comments. We also 
have not done any work in that area to look at it. I am not 
familiar with those tax practices necessarily, but it is an 
issue that does need to be considered, I think, as well.
    Mr. WELLER. Thank you. Thank you, Mr. Chairman.
    Chairman HOUGHTON. One of the things that I guess baffles 
me is that here we are. We are the Subcommittee on Oversight, 
and we can't get into the management here. We can ask 
questions. We can make suggestions. We can, I suppose, hold up 
money for the IRS, but that is a tough issue because it is such 
a vast operation, and in general, they are doing a great job.
    The Oversight Board that you are in charge of, isn't that 
part of management? You can make some recommendations, but you 
don't really get into the works. Carnegie Mellon and Mr. Dacey, 
you can do analysis. You can help, but you are not really in 
there.
    I would like to ask you, Mr. Cofoni, not trying to pin the 
tail on you or your excellent organization, at what time do you 
cancel the contract with a PRIME contractor? What prompts you 
to do that?
    Mr. COFONI. I would say certainly if we don't take the 
recommendations that have come from these studies that are root 
cause analysis and corrective actions, if we don't implement 
those actions and see improvement, then that would be a cause. 
If we don't deliver the end products of CADE and IFS, it seems 
to me those would be indicators. We are committed to doing all 
those things.
    Chairman HOUGHTON. At what point would you go to the 
Commissioner and say, it is impossible for us to operate here. 
The conditions are such that we cannot do the work which you 
expect of us and therefore we want to resign our contract.
    Mr. COFONI. We have begun to take positions--in the past--I 
want to correct a possible perception that we might take, 
continue doing work for revenue purposes. In all cases where we 
continued doing work before having a good set of requirements, 
it was with the best interests of the IRS at heart, people 
trying to be accommodating, trying to meet commitments, 
schedules that were necessary for internal performance or 
commitments externally.
    So--but we have had a bit of an epiphany. We really 
understand we can no longer do that, and we have put a stake in 
the ground and we have stopped work on projects when we thought 
we didn't have the right prerequisites to do a quality job. 
That is going to be the new pattern and that will be our 
behavior going forward.
    So, at what point, it is going to be at the point where we 
don't have clearly defined, detailed requirements, and it is 
okay. If we together don't understand what they are, then we 
should continue to explore, and if at the end of exploring we 
are still not convinced we know what they are, then we need to 
provide adequate reserve to deal with the unknown unknowns. So, 
I would say if we don't have detailed requirements, we are not 
going to go forward on any new work.
    Chairman HOUGHTON. When you deal with a consultant or with 
anyone from outside your shop and you buildup a body of 
knowledge, it is very difficult to cut the string because it is 
expensive, there is lack of education and personal contacts and 
things like that. Yet at the same time, it is important to look 
very firmly at whether people are doing the right job.
    Equally important as what a contractor has done, and we 
have touched on this all along, is what happens inside the 
shop. Mr. Levitan, you said that there is glacial progress as 
far as management getting a hold of this thing and squeezing 
it. Break it down a little bit, will you?
    Mr. LEVITAN. The specific area that I mentioned was 
improving the capability of the IRS to manage this program, and 
they just do not have people--the numbers of people or people 
with the depth of experience of managing programs of this scale 
to be able to do that. They are not capable of doing that at 
the present time. They have recognized it. They have initiated 
some searches. They have search firms working for them, trying 
to identify and then hire people that can bring that added 
competence. It has been moving very slowly and we have been 
urging them to make that move as quickly as possible.
    Chairman HOUGHTON. Let me just cut in here a minute. I see 
an almost impossible situation out there. If you are right and 
there is a lack of management ability to direct and enforce and 
monitor what is going on in terms of the PRIME contractor, and 
yet on the other hand in terms of the IRS internally there is a 
lack of confidence in what the PRIME contractor is doing, how 
do we get out of this mess?
    Mr. LEVITAN. That is the crux of the issue. Again, we get 
out of the mess by improving the capabilities of the IRS 
through process improvement and people and experience 
improvement, by working with the PRIME team, trying to make 
sure that we get the right people doing the right work that 
have the right skills, and that needs a lot more attention. It 
is going to take time, it is going to take a lot of work, and 
there will continue to be significant risks in that.
    Chairman HOUGHTON. We all know it ought to be done. The 
question is, will it be done and who does it and what part does 
this group play in this thing? We can ask questions. We can sit 
here, and we can be interested in your particular expertise. Is 
there a confidence that the next step will be taken so we are 
really going to get our hands around this thing so that next 
year when we have a session like this, a hearing, we are not 
going to be talking about the same issues? This has been going 
on a long, long time.
    Would any of the rest of you have any comments on that, Mr. 
Palmquist, Mr. Dacey?
    Mr. PALMQUIST. Mr. Chairman, one of the recommendations--
the fundamental recommendations we made in our report--was that 
at this point, there is no coherent systems engineering process 
on CADE which takes care of looking at CADE long term. The CADE 
has been allowed to focus on the short term because of the 
problems. To echo Mr. Levitan, the IRS itself does not have a 
dedicated systems engineer. They don't have a dedicated 
software architect or software engineer. They don't at this 
point have the technical staff to work with the program.
    The recommendation that we would have is to treat this 
improvement itself as a project along with the delivering of 
the deliverables. It needs to be measured, it needs to be 
agreed to by all stakeholders, so as you said, we don't arrive 
here a year from now. There have to be points along the way 
where you are looking and say, where are we? If this is not 
working, to have the defined actions in place that I will take 
step A or step B at this point, which, as you said, could 
include a change, could include a continuation.
    Chairman HOUGHTON. What chance would you have, before we 
move back to Mr. Dacey, what are the odds you would give to 
this thing working and getting back on track, 1 out of 10?
    Mr. PALMQUIST. As we told Commissioner Everson, we gave a 
60 percent chance that Release 1 would be delivered by the end 
of this year. That was Release 1 as defined to us back in the 
fall. I don't know if that functionality or design has changed. 
I understand that there are elements that have. I would 
probably -as an engineer, I am going to have to pick a 
deliverable--and so I will go with that. We predicted a 60 
percent chance of delivering CADE Release 1, which is 
fundamentally the infrastructure.
    The CADE Release 1 is actually viewed, in our opinion, 
incorrectly. It is far more complex than simply the small group 
of 1040EZs. When Release 1 is there, the infrastructure is 
there for the rest of the releases, but we went with 60 
percent, Mr. Chairman.
    Chairman HOUGHTON. How about you, Mr. Dacey?
    Mr. DACEY. In terms of the issues that we started to talk 
about, Mr. Levitan and Mr. Palmquist, there certainly is the 
human capital element of this which needs to go on. I think an 
important point that Mr. Palmquist made was you need to manage 
the process of fixing these issues almost as a separate 
project, and I think that is important. Again, one of the 
concerns we have had all along has been that the projects 
continue to roll on in the hope that some of these other things 
will get resolved, but there wasn't a separate process there to 
make sure they all got fixed.
    I would also like to say that in looking at this 
modernization process at IRS for quite a number of years now, 
it was only in the recent couple of years that some of the very 
foundational elements of system development were in place in 
enterprise architecture, a life cycle which specified how these 
processes were to take place and implementation of some of 
these improvements in the management practices. That is, again, 
a relatively short period of time.
    So, I think that those have been very positive steps in 
getting some of these foundational elements in place, but 
again, as we are finding in some of the testing, there are 
still risks that some of the things that weren't dealt with 
very early on in the program project development are creeping 
back in and causing problems today.
    I would like also to say that our relationship with the IRS 
has been positive. They have been open and candid and sharing 
very openly in our experience with these processes, which is a 
very positive step in my mind. They haven't tried to, in our 
mind, hide anything or do anything that would obfuscate our 
efforts. So, I think that is a very positive step in their 
acceptance of that. I don't know what the experience is of the 
other folks, but I assume it is similar.
    Chairman HOUGHTON. Okay. Mr. Pomeroy, do you have any 
further questions?
    Mr. POMEROY. I want to thank the panel. It has been a very 
interesting hearing, Mr. Chairman. I particularly have enjoyed 
Mr. Levitan's laying it right out there very straight. Having 
heard a lot of witnesses, I found your take on all of this to 
be very helpful.
    I also thought Mr. Palmquist's comments on managing the 
improvements as a separate project has an ongoing role for this 
Committee, Mr. Chairman. I would think we would probably want 
to reconvene this forum in a number of months, to be discussed 
in terms of what would be an appropriate time frame, but if we 
manage it as a separate project, I think there is an ongoing 
oversight role in seeing how we are coming. Thank you.
    Chairman HOUGHTON. Thank you. Gentlemen, thank you very 
much for your time and your wisdom and your advice. Thank you. 
Meeting adjourned.
    [Whereupon, at 11:20 a.m., the hearing was adjourned.]