[House Hearing, 110 Congress]
[From the U.S. Government Publishing Office]


 
   A REVIEW OF THE CENSUS BUREAU'S RISK MANAGEMENT ACTIVITIES FOR IT 
                              ACQUISITIONS

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

                                HEARING

                               before the

                  SUBCOMMITTEE ON INFORMATION POLICY,
                     CENSUS, AND NATIONAL ARCHIVES

                                 of the

                         COMMITTEE ON OVERSIGHT
                         AND GOVERNMENT REFORM

                        HOUSE OF REPRESENTATIVES

                       ONE HUNDRED TENTH CONGRESS

                             FIRST SESSION

                               __________

                           DECEMBER 11, 2007

                               __________

                           Serial No. 110-60

                               __________

Printed for the use of the Committee on Oversight and Government Reform


  Available via the World Wide Web: http://www.gpoaccess.gov/congress/
                               index.html
                     http://www.oversight.house.gov

                               ----------
                         U.S. GOVERNMENT PRINTING OFFICE 

43-197 PDF                      WASHINGTON : 2008 

For sale by the Superintendent of Documents, U.S. Government Printing 
Office Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800; 
DC area (202) 512-1800 Fax: (202) 512-2104 Mail: Stop IDCC, 
Washington, DC 20402-0001 




              COMMITTEE ON OVERSIGHT AND GOVERNMENT REFORM

                 HENRY A. WAXMAN, California, Chairman
TOM LANTOS, California               TOM DAVIS, Virginia
EDOLPHUS TOWNS, New York             DAN BURTON, Indiana
PAUL E. KANJORSKI, Pennsylvania      CHRISTOPHER SHAYS, Connecticut
CAROLYN B. MALONEY, New York         JOHN M. McHUGH, New York
ELIJAH E. CUMMINGS, Maryland         JOHN L. MICA, Florida
DENNIS J. KUCINICH, Ohio             MARK E. SOUDER, Indiana
DANNY K. DAVIS, Illinois             TODD RUSSELL PLATTS, Pennsylvania
JOHN F. TIERNEY, Massachusetts       CHRIS CANNON, Utah
WM. LACY CLAY, Missouri              JOHN J. DUNCAN, Jr., Tennessee
DIANE E. WATSON, California          MICHAEL R. TURNER, Ohio
STEPHEN F. LYNCH, Massachusetts      DARRELL E. ISSA, California
BRIAN HIGGINS, New York              KENNY MARCHANT, Texas
JOHN A. YARMUTH, Kentucky            LYNN A. WESTMORELAND, Georgia
BRUCE L. BRALEY, Iowa                PATRICK T. McHENRY, North Carolina
ELEANOR HOLMES NORTON, District of   VIRGINIA FOXX, North Carolina
    Columbia                         BRIAN P. BILBRAY, California
BETTY McCOLLUM, Minnesota            BILL SALI, Idaho
JIM COOPER, Tennessee                JIM JORDAN, Ohio
CHRIS VAN HOLLEN, Maryland
PAUL W. HODES, New Hampshire
CHRISTOPHER S. MURPHY, Connecticut
JOHN P. SARBANES, Maryland
PETER WELCH, Vermont

                     Phil Schiliro, Chief of Staff
                      Phil Barnett, Staff Director
                       Earley Green, Chief Clerk
                  David Marin, Minority Staff Director

   Subcommittee on Information Policy, Census, and National Archives

                   WM. LACY CLAY, Missouri, Chairman
PAUL E. KANJORSKI, Pennsylvania      MICHAEL R. TURNER, Ohio
CAROLYN B. MALONEY, New York         CHRIS CANNON, Utah
JOHN A. YARMUTH, Kentucky            BILL SALI, Idaho
PAUL W. HODES, New Hampshire
                      Tony Haywood, Staff Director


































                            C O N T E N T S

                              ----------                              
                                                                   Page
Hearing held on December 11, 2007................................     1
Statement of:
    Janey, Cheryl L., president of civil programs, Harris Corp.; 
      Judy Marks, president, Lockheed Martin Transportation and 
      Security Solutions; and Tom Romeo, Director, Federal 
      civilian agencies, IBM Global Business Services............    55
        Janey, Cheryl L..........................................    55
        Marks, Judy..............................................    60
        Romeo, Tom...............................................    69
    Kincannon, Charles Louis, Director, U.S. Bureau of the 
      Census; David Powner, Director, Information Technology 
      Management Issues, Government Accountability Office; and 
      Matthew Scire, Director, Strategic Issues, Government 
      Accountability Office......................................     6
        Kincannon, Charles Louis.................................     6
        Powner, David............................................    13
        Scire, Matthew...........................................    42
Letters, statements, etc., submitted for the record by:
    Clay, Hon. Wm. Lacy, a Representative in Congress from the 
      State of Missouri, prepared statement of...................     3
    Janey, Cheryl L., president of civil programs, Harris Corp., 
      prepared statement of......................................    57
    Kincannon, Charles Louis, Director, U.S. Bureau of the 
      Census, prepared statement of..............................     8
    Marks, Judy, president, Lockheed Martin Transportation and 
      Security Solutions, prepared statement of..................    62
    Powner, David, Director, Information Technology Management 
      Issues, Government Accountability Office, prepared 
      statement of...............................................    15
    Romeo, Tom, Director, Federal civilian agencies, IBM Global 
      Business Services, prepared statement of...................    71
    Scire, Matthew, Director, Strategic Issues, Government 
      Accountability Office, prepared statement of...............    15
    Turner, Hon. Michael R., a Representative in Congress from 
      the State of Ohio, prepared statement of...................    52


   A REVIEW OF THE CENSUS BUREAU'S RISK MANAGEMENT ACTIVITIES FOR IT 
                              ACQUISITIONS

                              ----------                              


                       TUESDAY, DECEMBER 11, 2007

                  House of Representatives,
   Subcommittee on Information Policy, Census, and 
                                 National Archives,
              Committee on Oversight and Government Reform,
                                                    Washington, DC.
    The subcommittee met, pursuant to notice, at 2:33 p.m., in 
room 2154, Rayburn House Office Building, Hon. Wm. Lacy Clay 
(chairman of the subcommittee) presiding.
    Present: Representatives Clay and Turner.
    Staff present: Darryl Piggee, staff director/counsel; Jean 
Gosa, clerk; Adam C. Bordes, professional staff member; 
Michelle Mitchell, legislative assistant, Office of Wm. Lacy 
Clay; and John Cuaderes, minority senior investigator.
    Mr. Clay. The Subcommittee on Information Policy, Census, 
and National Archives of the Committee on Oversight and 
Government Reform will now come to order. Today's hearing will 
examine the Census Bureau's planning and management of its key 
information technology systems and infrastructure to be used in 
the 2010 census. We will hear from the Census Bureau and GAO on 
their activity concerning the risk management of agency IT 
acquisitions for the upcoming census, as well as 
representatives of the key vendors involved with these 
projects.
    Without objection, the Chair and ranking minority member 
will have 5 minutes to make opening statements followed by 
opening statements not to exceed 3 minutes by any other Member 
who seeks recognition. Without objection, Members and witnesses 
may have 5 legislative days to submit a written statement for 
these materials for the record.
    I will begin with my opening statement.
    Welcome to today's hearing examining the Census Bureau's 
planning and management of key information technology systems 
and infrastructure for the 2010 census.
    In October of this year, the GAO issued a report entitled, 
``Information Technology: Census Bureau Needs to Improve Its 
Risk Management of the Decennial Systems.'' The study found 
that of the three acquisitions for the 2010 census, two were 
not on schedule and that the Bureau plans to delay testing 
certain functionality. As a result, GAO offered four 
recommendations for addressing the risk management problems 
related to IT for the 2010 census.
    Last month, the Department of Commerce Inspector General 
issued its semiannual report to Congress. During the review of 
this technology, the IG observed several problems: the handheld 
computer functions frequently froze; the processing of large 
address lists was slow; and help desk support for resolving 
users' computer problems was inadequate.
    The problems cited are urgent and must be addressed 
immediately.
    Today, we will examine the problems cited and 
recommendations offered by GAO and hear from the Census Bureau 
and the IT contractors for the 2010 census about their efforts 
to effectively and efficiently address the problems identified.
    And let me add to that, that this is not a dog and pony 
show. We are here for answers, and we want to hear what 
direction you will be taking as far as how we make this a 
complete and accurate census.
    [The prepared statement of Hon. Wm. Lacy Clay follows:]

[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
    
    Mr. Clay. Right now, as is the policy for this committee, 
we will swear in all witnesses, and I would ask you to stand 
and raise your right hands.
    [Witnesses sworn.]
    Mr. Clay. I ask that each of the witnesses give a brief 
summary of the testimony and to keep the summary under 5 
minutes in duration. Your complete written statements will be 
included in the hearing records, and Mr. Kincannon, welcome, 
and let us begin with you.

STATEMENTS OF CHARLES LOUIS KINCANNON, DIRECTOR, U.S. BUREAU OF 
  THE CENSUS; DAVID POWNER, DIRECTOR, INFORMATION TECHNOLOGY 
   MANAGEMENT ISSUES, GOVERNMENT ACCOUNTABILITY OFFICE; AND 
     MATTHEW SCIRE, DIRECTOR, STRATEGIC ISSUES, GOVERNMENT 
                     ACCOUNTABILITY OFFICE

              STATEMENT OF CHARLES LOUIS KINCANNON

    Mr. Kincannon. Mr. Chairman, thank you for the opportunity 
to be here this afternoon to discuss the key information 
technology systems under development for the reengineered short 
form 2010 census. Four major census IT systems illustrate the 
extent to which our Nation lies at the heart of our 2010 
operations: the MAF/TIGER Accuracy Improvement Project [MTAIP]; 
the Decennial Response Integration System [DRIS]; the Field 
Data Collection Automation program [FDCA]; and the Data Access 
and Dissemination System [DADS]. They are all critical to the 
success of the census, and I must say, the timing is critical 
to the success of these programs.
    Mr. Chairman, I cannot emphasize too strongly that we must 
have the necessary funding to carry forward these projects to 
success. And we must have that funding at the time that we are 
going to be able to use it to accomplish our tasks.
    As you know, the recent 7-week delay in funding census 
programs resulted from the first continuous resolution passed 
at the beginning of this fiscal year. This didn't allow a 
planned increase in census funding and forced us to delay and 
reduce the scope of our dress rehearsal. I want to thank the 
committee for its help in making sure that we got the funds 
that we needed, at least through Friday, and we will all hope 
for things to continue in the same vein.
    Over the next 3 years, delays in funding are one of the 
biggest risks that the 2010 census faces. Indeed, any 
additional delays will put greater risk in the face of the 
successful Census Bureau.
    Before I talk about what we are doing for 2010, I want to 
note that there is nothing new about the Census Bureau 
employing and developing new technology to improve the census. 
From the use of automatic tabulating machines from the 19th 
century to the development of the TIGER data base in the 1980's 
and our data capture system in 2000, we have been a pioneer in 
development for the use of technology to meet our needs. And 
the pattern of countries overseas following our lead in 
adopting the same technology has demonstrated that I think we 
are leading the way in a number of areas.
    From the 2010 census, the use of handheld computers 
represents the most fundamental change in census operations in 
many years, and they are the key to leveraging technology to 
improve the quality of census results and to control the costs.
    I want to emphasize to the committee that this is a new 
program for us. We have never done anything of this type on 
this scale before. Consequently, there are significant risks 
which are exacerbated by the strict time line that I mentioned 
earlier. It is possible that we will not have enough time to 
incorporate all of the functionalities that we have earlier 
planned. Adapting in this way is one way that we can reduce 
risks and still meet our schedule.
    I can report that the FDCA contractor, the Harris Corp., 
has provided a handheld computer that is functioning well in 
the initial dress rehearsal address-canvassing operations. The 
device has proved to be reliable, secure and user-friendly. We 
have successfully collected precise GPS coordinates for housing 
units and map features; data has been transmitted effectively 
both by LAN lines and by wireless technology. And our field 
workers are comfortable with the devices.
    As with any operation of this magnitude, the dress 
rehearsal is also identifying challenges. This is not 
unexpected. In fact, meeting these challenges is a fundamental 
step in the development of the 2010 systems and the very reason 
we conduct a dress rehearsal.
    Looking toward nonresponse followup, in the nonresponse 
followup operational test in the dress rehearsal next year, we 
will continue to monitor user problems. We will work with the 
Harris Corp. to assist handheld computer performance in terms 
of the fundamental objectives of the 2010 census.
    Our other contracts are on time and on budget. It is 
imperative that we test all of the interfaces between FDCA and 
our data capture system. After proving the functionality for 
nonresponse followup for the handhelds, this is the highest 
priority of our dress rehearsal.
    We have weathered the storm caused by the first CR, but 
just barely; the reason the GAO report on the status of census 
IT systems emphasized the need for an end-to-end systems test, 
both for systems supplied by contractors and those developed by 
the government. Because of the CR and the elimination of most 
of the paper-based operation originally planned for the dress 
rehearsal, there is an increased risk in the interfaces between 
these two sets of systems to mitigate the potential interface 
problems. We are considering additional testing in 2009.
    Your continued support is vital as we proceed with the 
development of the IT systems, and I thank you again for this 
opportunity to address these issues with you, and I look 
forward to your questions.
    [The prepared statement of Mr. Kincannon follows:]

[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
    
    Mr. Clay. Thank you, Mr. Kincannon, for your testimony.
    Our next witness will be David Powner. Please proceed.

                   STATEMENT OF DAVID POWNER

    Mr. Powner. Thank you, Mr. Chairman, for allowing me the 
opportunity to participate in today's hearing on the 2010 
decennial census information technology acquisitions. The use 
of automation will be critical to the success of the decennial. 
The Bureau estimates it will spend about $3 billion on 
information technology for the 2010 census. However, these 
technologies can present enormous risks and challenges if not 
managed effectively.
    With me today is Matt Stray, director of GAO Strategic 
Issues team, who has been closely monitoring the mobile 
computing device performance issues. Mr. Chairman, I would like 
to thank you and Ranking Member Turner for your early and 
frequent oversight of these acquisitions.
    In 2004, we started looking into the Bureau's institutional 
IT management capacity for you. We concluded from this review 
that there was much room for improvement. In March 2006, we 
testified before this subcommittee that neither the integration 
sytem nor the field data system collection project offices have 
the full set of capabilities needed to effectively manage these 
acquisitions.
    At that period, we stated that incomplete management 
activities, including those that required management, risk 
management and contract monitoring, increased the risk that the 
acquisitions will encounter problems in meeting costs and 
scheduled expectations.
    At this subcommittee's request, I will summarize our recent 
report on the status of four key acquisitions and discuss 
whether the Bureau is adequately managing key acquisition 
risks.
    In addition to the integration system in the field data 
collection system, which includes the mobile computing devices, 
there are two other major acquisitions, one to modernize data 
bases of addresses in maps and another to tabulate and 
disseminate data.
    The four acquisitions are showing mixed progress in meeting 
their costs and schedule estimates. The data base acquisition 
has been on schedule and within budget. The other acquisitions 
have been experiencing delays and one has experienced cost 
increases. Specifically, the dissemination contract has been 
awarded 2 years later than originally planned. The field data 
collection system cost estimate has increased several times due 
to poor cost estimation and requirements, and we project 
additional cost increases.
    In addition, both the field data collection system and the 
integration systems are deferring functionality to later bills, 
which typically results in the increased cost. In addition, 
deferring functionality means that the operational testing 
scheduled to occur during the dress rehearsal will not include 
the full compliment of decennial systems and their 
functionality. This raises the significance of systems testing 
post dress rehearsal.
    Given the relevant test plans were not completed, we 
recommended that the Bureau complete such plans, including end-
to-end testing to test the full compliment systems.
    Turning toward the management of the decennial 
acquisitions, the Bureau has identified mismanagement with its 
key acquisitions; for example, acknowledgement, which includes 
baselines, increasing requirements and aggressive test 
schedules. Despite this, we found three areas that could be 
strengthened: identifying risks, establishing mitigation plans 
and reporting those risks to key executives.
    For example, promoting mobile computer device performance 
issues associated with slow and inconsistent data transmissions 
had not been identified and tracked by the project office 
despite problems arising during the dress rehearsal. Because 
these devices are keystone to the reengineered census, it is 
essential that the Bureau perform the appropriate oversight of 
how the performance compares to what is expected and ensure 
that all performance limitations are figuratively addressed.
    We made a number of recommendations to the Bureau to 
approve the suspension activities, and to its credit, it is 
working on a national plan to strengthen these areas.
    In summary, Mr. Chairman, the IT acquisition plans for the 
2010 census will require continuous oversight. Although we are 
always seeing moderate cost increases, to date, the delay and 
functionality are a great concern because they will result in 
additional cost increases.
    These delays also elevate the importance of system 
integration in testing that will occur post dress rehearsal.
    Going forward, it is important that the Bureau closely 
monitor the cost schedule and function and delivery of its 
acquisitions; effectively manage its key risks associated with 
increasing requirements, system interfaces and mobile computing 
devices performance problems; and effectively plan and execute 
all systems testing, including the tests in the interrelated 
systems.
    This concludes my statement. Thank you for your leadership 
and oversight.
    [The prepared statement of Mr. Powner and Mr. Scire 
follows:]

[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]

    Mr. Clay. Thank you.
    Mr. Scire.

                   STATEMENT OF MATTHEW SCIRE

    Mr. Scire. Thank you for the opportunity to be here today. 
I believe the statement that Mr. Powner made reflects my 
thoughts as well.
    I would just add that we believe that continued oversight 
is very important at this critical point in time as we approach 
nonresponse followup, and look forward to working with the 
committee.
    Thank you.
    Mr. Clay. Thank you so much for that. And it is time to get 
to the questioning.
    Let me start with Mr. Kincannon.
    Director, we have seen delays in both the FDCA and DRIS 
systems acquisitions that have required the Bureau the 
establish later schedules for completing each project.
    How is the Bureau managing the risk associated with 
delaying key functionality for the 2010 census decennial IT 
system acquisition? And how will this impact the activities of 
the 2008 dress rehearsal?
    Mr. Kincannon. Mr. Chairman, the effects on the 2008 dress 
rehearsal are mixed with the CR and its effects on the dress 
rehearsal. But we believe that the slight delays in work on the 
DRIS contract, which is in fact within budget, will not 
adversely affect what we are doing here in the dress rehearsal.
    We had to make those rearrangements because, although our 
lifecycle estimates of the cost for the DRIS contract are still 
intact, the opinion and advice of the contractor was that the 
spacing by fiscal year was wrong. And we had to make some 
adaptations in order to shift funding for certain projects 
sooner. And therefore, other projects had to be delayed 
somewhat.
    The delays both from the CR and other aspects for FDCA do 
affect what will be tested in the dress rehearsal. But the key 
functionalities will be tested in the dress rehearsal. The 
integration system itself and the ability of handheld analysis 
to address a canvas has already been tested. The functionality 
was demonstrated. The problems were identified. And we will 
carry out later this spring the nonresponse followup for the 
dress rehearsal using the handheld, and then we will learn more 
about that test, I am sure.
    But we believe that functionality will go forward as 
planned starting in June.
    Mr. Clay. Mr. Powner.
    Mr. Powner. I think the key is, the functionality is not 
being tested going into the dress rehearsal. It needs to be 
tested after the dress rehearsal. That has been our point all 
along. When you look at the test plans that you would like to 
see in place, those currently aren't there. So the Census 
Bureau acknowledges those need to be done, and I think Mr. 
Kincannon mentioned that there is key 2009 testing that needs 
to occur, and it is important that we have a plan and we stick 
to that plan with the post dress rehearsal testing.
    Mr. Clay. How will the Census Bureau address that, Mr. 
Kincannon? Will you have back-up plans?
    Mr. Kincannon. We have plans to begin testing in 2009 to 
make sure that we cover all of the important functionalities.
    It is not a function of the handheld computers, but it was 
a function of DADS, which was delayed in being awarded, of not 
being able to test that with the dress rehearsal data. But we 
do plan to produce the data using the old DAD system, and we 
plan to test the new DAD system before 2010 with the data from 
the dress rehearsal passing through that system again and with 
data from the 2000 census to make sure that functionality is 
there.
    Mr. Clay. It is my understanding that the Bureau has 
engaged with the MITRE Corp. to evaluate the systems under 
development through a FDCA contract in order to test the 
reliability and effectiveness of the devices under development.
    Can you summarize your findings of MITRE's work for us, and 
would you also please submit all internal documents regarding 
MITRE's evaluation to the subcommittee for our records?
    Mr. Kincannon. The work that MITRE has done with us on a 
number of activities, not just FDCA but other activities in the 
planning for the 2010 census and for current activities of the 
Census Bureau, have been very helpful to us. We have not had a 
formal evaluation done by MITRE of the FDCA project, but they 
have reflected with us on certain activities, and we will be 
glad to provide those documents to the subcommittee.
    Mr. Clay. And what was their summary of FDCA?
    Mr. Kincannon. They have some concerns about the match 
between capacity to get the work done and the amount of time 
left to get the work done. And we are addressing that, and we 
will be continuing to address that both with the Harris Corp. 
and with MITRE and with our internal resources.
    It may lead, as I have said in my testimony, to determining 
that certain functionalities that were planned for the 
handhelds might be handled in a different way. And although we 
have not decided that, we are researching several areas to see 
whether there's a good way to handle those in an alternative 
way.
    Mr. Clay. Let me ask you this, Dr. Kincannon. Can you tell 
us with confidence that there are no inherent risks within the 
FDCA program that will require the Bureau to transition into 
contingency plans for a paper-based census?
    Mr. Kincannon. I don't see any--I don't see any major risk 
that we would have to transition into a full backup of a paper-
based census. I doubt that we have the resources to do that at 
this time, and I don't believe that it is necessary.
    I believe there will be functionalities where we have to 
choose different backup.
    I'll give you an example. In some hard-to-enumerate areas, 
there are, typically in every census, high rise buildings, 
private apartments or housing developments that have high 
nonresponse rates and require special action on our part. 
Optimally, we had planned that we would use the handhelds to do 
that special kind of operation which we referred to as a blitz. 
Maybe that's an exaggerated term, but it gets people busy.
    And, in fact, we have discovered them in the test in Queens 
where we were conducting a blitz in such a building with the 
same kinds of problems, that there were communications problems 
using the electronic devices that we had then.
    It is probably more practical in an isolated case like that 
to use a back up that is paper-based. You give everybody a 
stack of questionnaires. They start at the top of the packet 
and work their way down, or maybe go the other way around. And 
then you convert those paper forms, as we will for all of the 
mail-in forms, by scanning them into the system. That is an 
example of the way of--where we may well use a back-up system 
that is paper-based but not drop the automation plans now and 
try to plan a complete census based on paper.
    Mr. Clay. I would like to ask about the DADS two systems 
that will not be available for the 2008 dress rehearsal. What 
plans are in place to develop and test this system in time for 
the 2010 census?
    Mr. Kincannon. Well, the DADS system, the contract was 
recently just led on DADS to--as we referred to it, we have a 
lots of names for things, and they make me dizzy sometimes. 
When I am preparing for something like this, I have to ask 
repeatedly, now what does DAD mean, and what does, you know, so 
forth and so on. But you are experienced with this and not only 
at the census.
    We will have a system in time for the 2010 census, and we 
will have tested its functionality before 2010 by using the 
data from the 2008 dress rehearsal and also taking data from 
the 2000 census and running that big volume of data through 
DADS 2. So I think that is good.
    The same company that did DADS 1 won the contract for DADS 
2 in a pattern that is all too familiar, and I guess is our 
fate, so we are confident that they will be all to produce the 
updated system that is necessary because of the aging of 
equipment and methods used to do that delivery of data.
    Mr. Clay. In order to strengthen risk management activities 
for census acquisition, GAO made three recommendations to 
improve the process in place. These included an end to system 
testing, processes to mitigate risk and including senior Bureau 
leadership into decisionmaking activities.
    Please discuss your actions to address each of these 
recommendations.
    Mr. Kincannon. We are committed to end-to-end testing, and 
we have said that--what we are not able to test in the dress 
rehearsal, we plan to find the resources, find the time to do 
this in 2009 so we can be confident of all of those links 
between the different paper-based and electronically based 
systems and make sure about that functionality.
    On risk identification mitigation, we have, as the GAO 
observed, a number of provisions in effect and functioning in 
different offices, but they had very good suggestions for where 
we can strengthen that. I think it is basically true that we 
have agreed with their recommendations and are working to 
implement them.
    The involvement of top management in decisionmaking, there 
is pretty heavy involvement in the top management in the Census 
Bureau's top three layers of management in decisions about the 
operations and the procurement and the planning for the 2010 
census. And we would intend to strengthen that and make sure 
close attention is paid.
    I may have not done every bit of my duty here in my 
position because I have been in this very odd position, never 
sure whether October or November or December was going to bring 
release from my current responsibilities, but I'm going to 
assume now that I am going to be doing this for a while and 
will be paying close attention.
    Mr. Clay. You will be holding a place then.
    Mr. Kincannon. Yes, sir.
    Mr. Clay. One of the points that really stands out to me 
from the GAO assessment is that the risk management plans are 
pretty weak. And I want to know, is there any plans in place at 
this point to address the points of GAO brought up about risk 
management?
    Mr. Kincannon. Yes, sir. We transmitted to the GAO last 
week an action plan that provided information about how we are 
addressing that, and we have not heard back, but if they think 
we have missed some point, then I am sure we will hear further 
from them. And I am sure GAO can provide that to you. We can 
provide it.
    Mr. Clay. Let me find out from Mr. Powner.
    Have you had an opportunity to look at----
    Mr. Powner. We have looked. There does appear to be a 
commitment, as I mentioned in my oral statement. There is a 
commitment to putting national plans in place to more 
effectively manage risks, and that includes, clearly, three 
things: An acknowledgment of all of the risks. We saw some 
gaps. We think those gaps are closing. Having mitigation plans 
in place and having the key executives fully engaged in 
mitigating those risks going forward, and we have seen a 
commitment from the Bureau on that.
    Mr. Clay. This committee would be happy to get your 
assessment of the plan as well as what was submitted to GAO.
    Let me ask also Mr. Kincannon, the Bureau disagreed the 
GAO's recommendation with regard to performing end-to-end 
testing so that a full compliment of systems is tested in a 
census-like environment. In response, you have told GAO that 
you plan to test all critical systems and interfaces during the 
dress rehearsal and later. GAO tells us, however, the test 
plans are not complete.
    When will they be completed? And doesn't a decrease in the 
number of dress rehearsal operations increase the need for end-
to-end systems testing between the dress rehearsal and the 2010 
decennial?
    Mr. Kincannon. Our disagreement was--we weren't disagreeing 
with the principle, and we asserted we would be doing the end-
to-end testing in the dress rehearsal. Our commitment was based 
on the approval of the President's budget for the fiscal year, 
before the CR in other words. It was not desirable to eliminate 
some of the paper-based functionalities in it from the dress 
rehearsal test because it denied us the opportunity for an end-
to-end test in realistic census-like conditions. We cannot 
recreate fully those census-like conditions, but we can 
assemble those components and test them in 2009 in the event of 
a hot house kind of way to make sure the functionality is 
there. We are committed to try to do that.
    Mr. Clay. Mr. Powner, any response?
    Mr. Powner. I think it is fair to say, since our report was 
issued and we went back and forth on that issue, that the 
Bureau is clearly more committed to testing. That is our 
perspective on that, and I think a hearing such as this has 
helped with the situation here.
    Mr. Clay. OK.
    Mr. Kincannon, in 2005, GAO recommended that the Bureau 
define specific measurable requirements for the mobile 
computing devices and that they test the device's ability to 
meet those requirements in 2006.
    Again, in 2006, GAO recommended that the Bureau obtain 
validation and approval of baseline requirements for the FDCA 
project. Have requirements been developed? If not, then why 
not?
    Mr. Kincannon. Well, we now, from the address canvas, have 
some baseline data about the performance capabilities of the 
handhelds in field use. And we have discovered some problems 
and are dealing with those problems, but we do have a base of 
data about key aspects of their performance from which we can 
begin to develop standards that will define not only what we 
expect from the Harris Corp.'s devices but for what the 
productivity of individual and enumerators will be.
    And we agree with the GAO with whether we have to use the 
information derived from that activity as the basis for moving 
forward with practical, realistic goals, performance goals and 
measures so that we can set standards and then measure 
performances against those standards.
    Mr. Clay. Mr. Powner, any response? Or Mr. Scire.
    Mr. Scire. We have learned a little bit more about what the 
Bureau is doing in the area of measurement, and I think that we 
described it is that they have taken some first steps. We have 
learned that the Bureau wasn't even measuring average 
processing times, as an example. But as you know, an average 
disguises a lot so we would expect they could go beyond that, 
look at the distribution processing times, establish 
performance metrics that are expressed in terms of the percent 
of instances in which the handheld computers are meeting 
expectations.
    So, for example, you might have a performance metric which 
would say 95 percent of the time the handhelds are transmitting 
information within 12 seconds. That is just an example. I don't 
know that should be the exact number.
    But we would expect that the Bureau would then move in that 
direction and develop performance measures which are much more 
specific than simple averages. And that is a measurement of 
times. There are obviously other areas of performance of the 
handheld devices that you would also expect to develop 
performance measures that they could then use to hold Harris 
accountable for the work that they are doing.
    Mr. Clay. Any response to what Mr. Scire has said?
    Mr. Kincannon. Executive branch agencies complain about 
GAO, and they are always nagging about this thing or the other. 
But GAO has been quite helpful in this case in pointing out 
reasonable things that we need to do that will help make for a 
better census. And I think we are going to profit from that.
    Mr. Clay. Let me ask Mr. Scire or Mr. Powner. Please 
describe for us the major flaws inherent in the Bureau's risk 
management strategies for the decennial IT acquisition. Are the 
flaws based upon lacking or ill-defined system requirements 
during the design phase, or are there other contract management 
issues that contribute to the problem?
    Mr. Powner. Clearly, if you look at their risk management 
activities, some of the things you mentioned there, 
requirements, management and contract oversight, those are a 
couple of key risks and that is nothing new. That is something 
that we have been reporting and you have been asking questions 
for several years now on.
    When you look at their risk management activities, again, 
what we saw was first of all, certain risks were clearly being 
made as part of their formal risk management program. For 
example, system interfaces between the systems, that seemed to 
fall between the cracks, having the appropriate mitigation 
plans in place, and also we were looking for key evidence that 
the executives were engaged in mitigation of those risks.
    So those were the key areas where they were lacking.
    Mr. Clay. And you think they have begun to address them?
    Mr. Powner. Yes. I mentioned the action plan that they sent 
over to us just recently. That is a good start in the right 
direction to more appropriately manage these risks, but going 
forward, there is a lot of work because some of these are going 
to be around for a while. Especially when you start looking at 
the requirements, creating the remaining testing and pushing a 
lot out into the later bill to try to get more development 
testing done in the later phases, and that is difficult given 
the moveable deadline.
    Mr. Clay. GAO reported that the FDCA cost estimate has 
increased by more than $50 million and that additional cost 
increases are expected. What are two reasons for the increase, 
and are the cost increases correlated with deficiencies in the 
designs or incomplete definitions of system requirements and in 
the contracts agreed to with the vendors?
    Mr. Powner. If you look at the cost increases to date, Mr. 
Chairman, clearly incomplete requirements and growing 
requirements is one reason why we see increase in costs. 
Another key reason was a poor estimate to begin with.
    Mr. Clay. So are you saying the Census Bureau did not 
exactly know what they were purchasing?
    Mr. Powner. I think they knew what they were purchasing, 
but when you have incomplete requirement definition up front--
and the Census Bureau isn't alone in this. We see this commonly 
throughout the Federal Government where you have incomplete or 
not a complete cost estimate to begin with. I mean, we had a 
contract that we increased contract costs twice already. We 
actually have a technique where we look at burn rates and 
project overruns. We project additional ones going forward, and 
I think, with growing requirements, we will expect more 
increases.
    Mr. Clay. Mr. Powner, for the viewing public, break that 
down into I guess household terms. If we were purchasing 
something for a household, give me an example of what went 
wrong here with the $50 million overrun. How would we----
    Mr. Powner. Clearly, if you look at the reasons for the 
overrun, it was increasing requirements and, of course, cost 
estimates. So if you were building a house, you would have 
aspects of your house, you know, in terms of square footage, 
you know, the features you have learned in your kitchen and 
those types of things; you would define those features going 
forward. That is what you would expect in a border. So it is no 
different from a system.
    With your system requirements, you would want to see 
specificity in terms of exactly what you want so that the 
contractor can then carry out that plan.
    As you start adding requirements to a system, it is the 
same thing as when you start adding systems to your home. If 
you want something more in the kitchen and want additional 
square footage or this feature or that feature, you are going 
to start seeing the cost go up, and that is exactly what is 
happening with that system.
    Mr. Clay. So if we wanted marble countertops, that would 
add a little bit more to it.
    All right. Thank you for that explanation.
    Let me also ask you, GAO recommended that the Bureau 
perform end-to-end testing on its system. Why is this so 
important, and what are your concerns in the Bureau's plans in 
this area? Are the Bureau's reasons for resisting this idea 
reasonable?
    Mr. Powner. Today we are hearing that there is a 
receptivity to the end testing, which is a good thing.
    The important item here is, because not everything will be 
tested during the dress rehearsal as originally planned, the 
inter relationships of these many systems, and there are many--
we talked about four major acquisitions today. There are legacy 
systems, and there are a lot of interfaces here that need to 
work. So it is important that we have the appropriate 
integration testing and testing to make sure that not only the 
individual pieces work, but they work as a whole.
    What we did not see was the test plans in place to make 
sure that this happens. There is a commitment to do the end-to-
end testing now, we are hearing, in 2009. And that is a good 
thing. But that will also require continued oversight to make 
sure those test plans are complete and that they are vigorously 
executed.
    Mr. Scire. If I can add to that, the importance I think is 
that sort of testing be done under census-like conditions. As 
Mr. Kincannon was saying, that is where you are going to see 
the limitations of the systems. And for the nonresponse 
followup to the dress rehearsal, there's a critical interface 
here that needs to be tested then. It is not something that it 
could be tested later, and that is the interface between FDCA 
and DRIS, and how that works with late returns.
    One of the arguments for introducing the handheld was a 
cost savings that would accrue by doing this late mail return. 
That gives you that capacity. So that is something where that 
interface would be important to be tested during the dress 
rehearsal rather than later.
    Mr. Clay. So those are some of the areas that this 
subcommittee should continue to have oversight over.
    Mr. Scire. Absolutely.
    Mr. Clay. And leading up to the 2010.
    So any other areas we need to possibly exercise oversight 
and really pay attention to?
    Mr. Powner. I think if you look at the testing going 
forward, that is a key one. In monitoring the cost and 
scheduling performance of these major acquisitions, clearly you 
want to do that and then also to--the performance and 
resolution of the issues with the mobile computing devices, 
that would be a third.
    Mr. Scire. If I could add to that, as far as looking at the 
computing device. I think it is true that we still don't know 
the magnitude of the performance issues that we and the IG and 
the census itself observed during the address-canvassing dress 
rehearsal. So I think it is something that deserves continuing 
oversight.
    Mr. Clay. And in your opinion, the top three acquisition 
risks facing the Bureau between now and the 2010 census?
    Mr. Powner. I would say the increasing requirements, 
managing the many interfaces and the remaining testing.
    Mr. Clay. Mr. Kincannon, anything else to add?
    Mr. Kincannon. I had a long dry spell there, Mr. Chairman. 
But I have two or three.
    Mr. Clay. We will always give you the opportunity to 
respond.
    Mr. Kincannon. It is true that we think also just about the 
most important thing, well, the functionalities of the handheld 
in the dress rehearsal will be tested, and that very important 
one of how we deal with late rural returns is a big money 
saver. In the test censuses in 2006, up to 14 percent of the 
receipts were late mail receipts, and that translates into a 
lot of savings if we get that information immediately 
transmitted back to the handhelds in the field so we don't send 
people to knock on those doors.
    In the old system with paper, we were never able to catch 
up so we would have to knock on those doors again, those being 
irritated a second time, and we have a second piece of paper 
and sent it in and then it had to be duplicated. So that is 
very important, but we are planning to test that in the dress 
rehearsal. We endeavored in making our modifications to the 
dress rehearsal, as we deleted or constrained things under the 
DRIS--under the CR, we tried to preserve the most important 
things that we really have to have good knowledge about the 
functionality. We would have liked to have had it all, but we 
couldn't have it all.
    We have mistakes and errors that we have made, but some of 
them would have been avoided if we hadn't gotten--by the CR.
    Let me also say that, of these four contracts, three are 
essentially on schedule. We have made some schedule 
modifications and within budget, but one of them is only just 
beginning. So that is not a fair test, but all three of those 
have, as a characteristic, they are things that we have done 
before with contractors. And so those have worked very well 
because we had experience with them.
    The problem with the FDCA was it was something that we have 
not done before, and we did not do a good job of understanding 
what the cost should be.
    And so we did have to make a change both in the overall 
cost and in the timing by fiscal year of funding this meeting.
    I think that is a distinguishing characteristic.
    The GAO report has some very handy little charts, sort of 
like the consumer reports chart. You know you are going to buy 
a car. They are the same thing. A little empty circle means you 
are not doing it right, and a full circle means you are doing 
it right. And I think you file consistently the processing 
order, so that No. 1 is intake and so forth. Winding up with 
four being DADS. They have been very logical. The intake really 
looks like the worst of it. It has the most half circles and a 
couple of completely empty circles, and yet it is on time and 
within budget. And that is not because we shouldn't be excused 
of doing these things, but it is because we understood the 
process and exercised good control even without following some 
of the proper procedures. But that makes it very important that 
we follow the proper procedures on FDCA.
    Mr. Clay. And you know, Mr. Director, you have with your 
tenure here in Washington, with your service at the Bureau and 
with, I guess we would put it as your tentative stay at the 
top, and we will get a successor for you; you have been through 
this before. You know you cannot count on a CR that--we don't 
know if you get a CR, you get an appropriations bill, and you 
understand the work of Washington, and that is why it is so 
important that we get this right. And yes, there will be a 
dress rehearsal in 2008 but you don't get a dress rehearsal in 
2010, and we need to get it right. And I know you are aware of 
that, and under your stewardship, just keep us on track for 
2010 census.
    Mr. Kincannon. I will do my best, sir, thank you.
    Mr. Clay. I know you will.
    And with that, I will dismiss this panel and call up the 
second panel. Thank you all for your statements and testimony.
    On our second panel, we have a highly distinguished group 
of individuals who are highly qualified to address the issues 
associated with the four major IT acquisitions underway for the 
upcoming decennial census.
    And beginning to my left is Ms. Cheryl L. Janey, who is the 
president of the civil programs business unit of the Government 
Communications System Division of Harris Corp. There she 
overseas the development and production of advanced 
communication systems for agencies of the U.S. Government and 
their prime contractors.
    And welcome, Ms. Janey.
    Ms. Judy F. Marks, is president of Lockheed Martin 
Transportation and Security Solutions, A division of the 
Lockheed Martin Corp. In this role, she manages three lines of 
business which focus on advanced mission, critical information 
technology solutions, including Census Data Capture and 
Communications Netware Infrastructure Program.
    Thank you for being here, too.
    And Mr. Tom Romeo serves as the director of Federal 
civilian agencies for IBM Global Business System Services. In 
this role, he is responsible for all IBM services, business 
relationships and contracts throughout the Federal, Civilian 
agency community, including the Department of Commerce and 
Census Bureau.
    And I welcome you all together.
    And it is the policy of the subcommittee to swear in all 
witnesses before they testify. I ask you to stand and raise 
your right hands.
    [Witnesses sworn.]
    Mr. Clay. Thank you. Let the record reflect that all 
witnesses answered in the affirmative.
    I will ask each witness to now give an oral summary of his 
or her testimony and keep it under 5 minutes in duration. Bear 
in mind that your complete written statement will be included 
in the hearing record.
    And without objection, I would like to submit the opening 
statement of my colleague and ranking member, Mr. Turner of 
Ohio.
    [The prepared statement of Hon. Michael R. Turner follows:]

[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
    
    Mr. Clay. And we will now begin with Ms. Janey.
    You may begin.

  STATEMENTS OF CHERYL L. JANEY, PRESIDENT OF CIVIL PROGRAMS, 
     HARRIS CORP.; JUDY MARKS, PRESIDENT, LOCKHEED MARTIN 
TRANSPORTATION AND SECURITY SOLUTIONS; AND TOM ROMEO, DIRECTOR, 
    FEDERAL CIVILIAN AGENCIES, IBM GLOBAL BUSINESS SERVICES

                  STATEMENT OF CHERYL L. JANEY

    Ms. Janey. Congressman Clay, my name is Cheryl Janey, and I 
am the president of the civil business division for Harris 
Corp. I am pleased to have the opportunity to discuss the role 
of Harris in supporting the Census Bureau in the 2010 
decennial.
    The Field Data Collection Automation [FDCA], program was 
awarded to Harris in April 2006. Since the contract was 
awarded, we formed a team of highly skilled professionals 
focused on successfully supporting the 2010 census. We are 
delighted with the progress to date and are proceeding at an 
aggressive pace.
    The FDCA program provides the automation support for the 
Bureau to collect quality data in an efficient and cost-
effective manner for the 2010 census. This includes the 
hardware, applications and infrastructure necessary to support 
field activities.
    We interact daily with the Bureau to manage the technical 
schedule and cost risks of the program. Rigorous testing has 
been conducted and will continue throughout the FDCA program.
    The handheld device marks the first time enumerators will 
use electronic means to collect and record data. This is a 
historic milestone for the Bureau and one which must be met 
with careful planning and testing to ensure the data remains 
secure, the process efficient, and ultimately that the 
decennial is accurate and complete.
    In spring 2007, Harris delivered a secure, robust and 
reliable system as part of the dress rehearsal address-
canvassing field operation. During address canvassing, we 
successfully deployed nearly 1,400 intuitive handheld devices 
developed by Harris. Key FDCA supported infrastructure were 
deployed, including the Network Operations Center, Security 
Operation Center, Data Processing Centers, and a help desk. 
Overall, the testing and the handheld reliability exhibited 
during the spring DRAC field operations was encouraging. 
Valuable information was gathered through the process, which 
was the purpose of this early field evaluation.
    When necessary, Harris utilized secure over-the-air 
software upgrade procedures to correct defects and maintain 
operational effectiveness. Some challenges surfaced, including 
issues with transmission speed and synchronization, but this is 
understandable at this phase of a program of this size and 
complexity. Harris developed temporary fixes to the problems 
encountered and is actively working toward permanent resolution 
in time for the planned operational tests.
    Using the systems engineering approach, we established a 
lessons learned review board. This board prioritizes and 
reviews corrective action plans, including the testing process. 
Once fixes are made and tested, they are integrated into the 
system and the system test is run to ensure they work to 
accomplish the desired results. We have followed this process 
with critical improvements to transition, speed and 
synchronization time, among others, ensuring they perform as 
designed in the upcoming operational tests.
    The security of the collected data has been a paramount 
concern to the Bureau and also of Harris. Multiple overlapping 
layers of security have been embedded in design and deployment 
of the handheld devices. We have created security systems to 
protect Title 13 and other sensitive data during collection and 
transmission and at any point throughout the process.
    The Bureau recently commissioned an independent assessment 
of the security measures. This assessment validated the 
technical and procedure designs and risk mitigations that we 
have incorporated into the program to safeguard data.
    Given the unbending census date of April 1, 2010, we have 
limited time to incorporate any changes required as a result of 
field integration and field testing. The recent period of 
reduced funding during the first Continuing Resolution did have 
some impact on timing and the scope of the planned NRFU dress 
rehearsal. Harris is actively working with the Bureau 
developing a revised testing approach for NRFU and all 
remaining operations that will meet both the financial and 
timing limitations of the Bureau.
    We have confidence in the capability and performance of the 
infrastructure and are moving carefully and thoughtfully 
through the planning process to ensure reliability is not 
compromised and integrity is maintained.
    Harris Corp. will continue to support the Bureau in 
managing risks and will contribute in any way to make sure the 
2010 decennial will provide the most accurate, complete and 
secure count of our Nation's population.
    Mr. Chairman, I appreciate the opportunity to testify 
before you today. I look forward to answering any questions you 
may have.
    [The prepared statement of Ms. Janey follows:]

[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
    
    Mr. Clay. Thank you so much.
    Ms. Marks, you may proceed.

                    STATEMENT OF JUDY MARKS

    Ms. Marks. Thank you, Chairman Clay.
    My name is Judy Marks. I am president of Lockheed Martin 
Transportation and Security Solutions. I appreciate the 
opportunity to speak to this community. I am pleased to share 
the progress we are making on the Decennial Response 
Integration System [DRIS], program for the 2010 census. Today I 
will describe what the Lockheed Martin team has done and is 
doing to keep the DRIS program pivotal to the success of 2010 
on track.
    First, I want to reassure this committee that we understand 
how important the census is to our Nation. The constitutional 
mandate is a weighty responsibility. Certainly one such as the 
census merits special respect and consideration for the results 
touch every one of us in this room, indeed touch every 
individual American.
    To this end, I am pleased to report that, to date, 100 
percent of the DRIS program deliverables have been made on time 
and in full compliance with the requirements. Together, 
Lockheed Martin and the Census Bureau are on schedule and 
within budget for this core system.
    Our team supported the census in conducting the 2000 
census, the most accurate in our Nation's history, and we have 
effectively applied lessons learned from our 2000 experience. 
Lockheed Martin's leadership program began in 2005 following a 
competitive proposal process. The responsibilities of the DRIS 
program include designing, building, testing, deploying, 
implementing, operating maintaining, and securing the systems, 
infrastructure, staffing, procedures and all of the facilities 
needed to successfully carry out the 2010 census.
    Through these activities, we provide assistance to the 
public through the telephone. We will receive, capture and 
standardize census data provided to telephone agents or through 
census forms, and we will receive standardized data collected 
by the handheld computers.
    Following the conclusion of the census activity, we will 
also dispose of the systems and infrastructure associated with 
the census, and finally, we will decommission the 2010 
facilities and staff.
    Lockheed Martin has remained within the original Census 
Bureau total lifecycle funding in addressing DRIS requirements, 
and we remain committed to delivering DRIS solutions within the 
planned lifecycle funding and on schedule.
    Our role is distinctly separate from two other components 
of 2010 census represented on this panel, the FDCA program by 
the Harris Corp. and the Data Access in Dissemination Program 
led, too, by IBM.
    I would like to now touch on some of the highlights of the 
DRIS program to date.
    First, we have completed system development for the 
upcoming dress rehearsal system and are now in the midst of 
system integration test efforts. This system has been deployed 
and is currently being tested and certified at the national 
processing center in Indiana.
    And all of these activities prepare our team for the dress 
rehearsal in May 2008 where we will test the solution and 
identify areas that still require refinement prior to 2010. We 
have already demonstrated multiple functions of the 2008 dress 
rehearsal system to the Census Bureau and to other 
stakeholders, thereby continuing to reduce risk to this test. 
In the time that remains before 2010, Lockheed Martin and the 
Census Bureau will focus on the following: We will continue to 
implement a comprehensive system test approach which will drive 
performance, which will enhance quality and which will reduce 
risk. We're actively engaging in services of small businesses 
that can add value in the DRIS program.
    Currently, our small business participation objective is 30 
percent of our contract value, and I'm proud to say we're on 
target to surpass this objective. We're continuing our proven 
record of earned value management, scheduling management and 
risk management on the program to ensure that DRIS remains 
fully compliant as it is today and we'll continue to operate as 
an integrated highly cooperative government industry team from 
which we all benefit. The census is absolutely critical to 
every American citizen.
    The data the Bureau collects during the process helps 
foster our democratic process. In order to achieve the success, 
the Census Bureau must rely on support from an array of people, 
processes and technology. The DRIS program will use information 
technology and automation to accurately securely and 
efficiently count this Nation's population. We are 
accomplishing this by advancing a strong foundation we've built 
in partnership with the Bureau, a collaborative team structure, 
proven risk reduction and program management practices, focus 
on inclusion of small businesses, and we continue to deliver 
the right capability on time and within budget. At Lockheed 
Martin, we are committed to serving the U.S. Census Bureau with 
excellence and partnership to carry out this critical 
congressional mandate in 2010.
    Mr. Chairman, I look forward to your questions on this 
statement and my written testimony. Thank you.
    Mr. Clay. Thank you so much, Ms. Marks.
    [The prepared statement of Ms. Marks follows:]

[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
    
    Mr. Clay. Mr. Romeo, you may proceed.

                     STATEMENT OF TOM ROMEO

    Mr. Romeo. Thank you, Mr. Chairman. And thank you for the 
opportunity to testify before you today. My name is Tom Romeo. 
I'm the director of Federal services for IBM's Global Services 
Business in the public sector. I'm here today to talk about 
IBM's role in supporting the Census Bureau for the 2010 U.S. 
census, especially focussing on those concerns identified in 
the GAO report on October 2007. IBM has a long history of 
working with the Census Bureau. The first automated census of 
1890 was the inspiration for the birth of the Hollerith card, 
the foundation of modern computing, which remained in use 
through the 1970's. Herman Hollerith's company was one of the 
founding companies of the IBM Corp. In more recent times, IBM 
supported the 2000 census as the prime contractor for the first 
Data Access and Dissemination System [DADS] contract, providing 
both data tabulation and Internet data dissemination.
    In 2005, IBM was proud to be part of the winning Lockheed 
Martin team on the 2010 Decennial Response and Integration 
System [DRIS]. Our role in that contract was to provide the 
systems supporting both the telephony and Internet data 
collection channels for the 2010 census. In September of this 
year we were awarded the DADS II contract, and will again be 
providing data tabulation and Internet data dissemination 
services for the 2010 census and for other Census Bureau 
surveys.
    The October GAO report identified various concerns 
regarding the schedule and status of the programs with which we 
are involved. And we would like to comment briefly on these. 
With respect to the DADS II contract, although an earlier award 
would have allowed us to begin development sooner, we do not 
believe the delay is a significant risk to the timely 
tabulation of the 2010 census data. We should point out that 
the original DADS contract was awarded in April 1997, only a 
few months earlier in the decade than the new DADS II contract.
    At that time there were no existing tabulation or 
dissemination systems, so the risks were arguably higher than 
it is today. In addition, the proposed replacement tabulation 
system is built on the same technology and architecture as the 
original tabulation system, so the upgrades required to make it 
ready for the 2010 census are not as significant as was 
required to build the original system for the 2000 census. 
Using the current tabulation system to support the 2008 dress 
rehearsal, although not ideal, is a completely workable and 
low-risk approach to meeting current schedule constraints.
    With respect to data dissemination, our system development 
schedule is built around the launch of the new system in early 
2011, and we believe the schedule will give us sufficient time 
to achieve our objectives. The GAO report also mentioned that 
the DRIS Telephone Questionnaire Assistance capability, that is 
the inbound calling functions, will not be developed in time to 
support the 2008 dress rehearsal. The funding constraints from 
fiscal year 2006 through fiscal year 2008 described in the GAO 
report did require the exclusion of some telephone system 
functionality.
    However, most of the functionality selected for exclusion 
from the 2008 system was part of the 2000 census, and was 
therefore a lower risk for later deployment. The dress 
rehearsal telephony system focused instead on outbound calling 
functions that were not implemented for the 2000 census. We do 
support additional possibly end-to-end system testing in 2009 
that includes the full set of telephony features, which is what 
the Census Bureau currently plans.
    In closing, we would like to express both our commitment to 
seeing the Census Bureau through a successful 2010 census and 
our appreciation for the Census Bureau's work today. In our 
long history of working with the Census Bureau, we have been 
thoroughly impressed by their professionalism and dedication of 
both their employees and leaders and by their focus on 
continuous improvement in technology innovation.
    Thank you for the opportunity to testify and I'm happy to 
answer any questions.
    Mr. Clay. Thank you so much, Mr. Romeo. And thank you all 
for your testimony.
    [The prepared statement of Mr. Romeo follows:]

[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
    Mr. Clay. Let me start with Ms. Janey.
    Significant concerns have been raised that Harris is 
scheduled for deliverables, such as software and hardware, may 
not meet the Bureau's schedules for deliverables under FDCA. 
Can you assure us that your schedules are in concert with the 
Bureau's deadlines and needs?
    Ms. Janey. I can, Mr. Chairman. We are working with the 
Bureau on a literally daily basis as well as with the GAO to 
ensure that our delivery schedule matches the means needs of 
the Bureau.
    Mr. Clay. OK. Let me ask a question similar to what I asked 
Mr. Kincannon earlier. Can you describe for us the role Harris 
played in the FDCA evaluation conducted by MITRE Corp? What 
types of information or data did Harris provide to MITRE for 
the evaluation? Can you summarize the findings of MITRE and its 
characterization of Harris's work under the FDCA contract? And 
can you state with confidence that there are no interests--
inherent risk within the FDCA program that will require the 
Bureau to transition into contingency plans for a paper-based 
census?
    Ms. Janey. Well, I will start with the end of that first. 
Any time a new system is implemented, it's a challenge and 
there are risks to it. That's why Harris, in conjunction with 
the Bureau and the varying oversight agencies that are working 
with the Bureau, are focusing so keenly on ensuring that we are 
sticking to a plan and sticking to a schedule. I can't speak 
specifically about the findings that MITRE gave. I think I 
would direct you to the Bureau or to MITRE themselves. Harris 
regularly provides significant amounts of data both at the raw 
data summary level and everything in between. We did coming out 
of the dress rehearsal where there were some synchronization 
challenges and timeframes, and have since provided updated 
information back again to the Bureau and to GAO. So I can't 
speak specifically for the MITRE summary, but I can tell you 
that Harris has provided any information that's requested and--
--
    Mr. Clay. Has MITRE responded back to you all--to Harris 
with a summary?
    Ms. Janey. No. I think MITRE was working with the Bureau.
    Mr. Clay. I see. OK. Thank you for that response. Let me go 
to Ms. Marks. Apparently, Ms. Marks, the DRIS project has 
already experienced one scheduling delay and has been altered 
to operate at a reduced level of functionality. Can you explain 
why this is? And was it solely due to inadequate system 
requirements, definitions from the Bureau?
    Ms. Marks. Mr. Chairman, thank you for that opportunity to 
answer. The challenge that DRIS ran into when we were 
competitively selected in 2005 is the Bureau had done their 
best to identify all of the program requirements that they 
could that they knew at the time. For example, there are 
multiple forms used in the census.
    For the purpose of running a competitive procurement, the 
Bureau selected one representative form to have both ourselves 
and the competition bid. It turns out today there are 62 unique 
forms and they continue to be defined as we go into 2010. All 
of those forms will be tested on a paper basis at the dress 
rehearsal. So it's those kind of additional finite definition 
that happen as you continue to go through the decade and as you 
get closer to the census that are to be expected.
    So we are staying within the life cycle limit. We have 
stayed within that funding profile and we do look forward to a 
successful 2010 census because of that.
    Mr. Clay. Thank you for that response, Ms. Marks, because 
of the delays in DRIS, the Bureau will not have a telephone 
questionnaire assistance system in place for the dress 
rehearsal. Normally they would have these data capture centers 
complete by the end. How will you seek to mitigate future 
system vulnerabilities that arise between the dress rehearsal 
and the actual 2010 decennial census?
    Ms. Marks. Well, Mr. Chairman, as we were defining what 
would go into the dress rehearsal with obviously some of the 
changes that were occurring, we prioritized functions in the 
following manner: If we had a function that worked--and we are 
very proud to have been the 2000 census provider--if there was 
a function that worked, like the telephone questionnaire, we 
prioritized that to be tested at a later date. What we wanted 
to test early were the functions that had never been in use 
before. The most important function being the interface face 
with FDCA. We are testing all of the primary interfaces with 
FDCA at the 2008 dress rehearsal, and we believe that is the 
most critical risk item to retire between ourselves, the Harris 
Corp., and the Bureau. All of the other functions, they are not 
going to be in dress rehearsal in 2008. We have proposed, 
again, within that life cycle funding to the Census Bureau to 
do it in 2009, including the telephone questionnaire.
    Mr. Clay. And you are pretty comfortable with the telephone 
questionnaire?
    Ms. Marks. We are. It worked successfully in the past.
    Mr. Clay. Let me go to Mr. Romeo. And thank you for your 
response. How will the late development of DADS II affect your 
ability to ensure that the system will be adequately integrated 
and tested in time for the decennial? What challenges do you 
foresee that what may require further scheduling delays or cost 
overruns? And can you describe for us how you plan to test the 
full functionality of DADS II while it is in development?
    Mr. Romeo. Sure. Thank you, Mr. Chairman. So the award was 
a little bit later than we had hoped. So we didn't get started 
as soon as we had hoped. But the testing today--the plan for 
the testing is to use the existing DADS system for the dress 
rehearsal test and then to retest with the newly developed 
system. The data in the DADS system is the data collected by 
DRIS passed to the Census Bureau cleanse and then passed to 
DADS for analysis and presentation to the public. And because 
of that schedule, it is a later requirement in the system. The 
interfaces between DRIS and the DADS system are very similar to 
the interfaces that we implemented for the 2000 census, and 
they are fairly limited. There's two interfaces. So we're very 
confident that the test, using the existing system, will give 
us a great head start and the later test with the new system 
will be adequate to ensure the functionality.
    Mr. Clay. Thank you. Let me ask a panel-wide question. 
We'll start with Ms. Janey and just move down the line.
    Because the Bureau has delayed its schedule for FDCA, DRIS 
and DADS II contract as well as delayed functionality of key 
system activities beyond the dress rehearsal, there are 
increased risks associated with system integration and 
interoperability among all four acquisitions. Can each of you 
please describe for us how you are mitigating the risk 
associated with system compatibility and interoperability prior 
to April 2010, and has the Bureau effectively managed its 
enterprise architecture development activities to ensure its 
systems are fully interoperable and they're integrated? I'll 
start with you, Ms. Janey and see if you can tackle that.
    Ms. Janey. Well, I liken it to a relay, Chairman Clay. Each 
individual runner in a relay can operate at his personal best, 
but that relay team won't win unless the handoffs are 
efficient. I think the same can be true of the criticality of 
the interfaces between DRIS and FDCA particularly in this 
census. I'm happy to tell you that Harris is working with the 
Bureau, with Lockheed Martin to ensure that we test and 
rigorously test those interfaces as they developed. Is there as 
much time as we'd want? No. But I don't think there is ever as 
much time as we want. We are developing rigorous testing plans 
at the Bureau's direction and in cooperation with the Bureau, 
with Lockheed Martin and with all of the contractors involved 
to ensure that we've tested it adequately far before the 2010 
census.
    Mr. Clay. OK. Thank you for that. Ms. Marks.
    Ms. Marks. Mr. Chairman, the DRIS system accepts data from 
three sources. We either receive the paper forms, we receive 
inputs via the telephone VRE call centers or we receive them 
electronically from the FDCA system. And then as Mr. Romeo 
shared, some of that data after we submit all of this data to 
the Census Bureau, they are the only people who cleanse it and 
then several of--some of that data then goes to the DADS 
system.
    So we have the ability to accept data in any one of three 
manners as a secure manner and we test each of those 
rigorously. We have already started testing prior to dress 
rehearsal some files coming to from FDCA to get basically an 
advanced start on testing some of the interfaces. Again, all of 
those primary interfaces will be tested in 2008, and we always 
have the ability to continue in 2009 in the end-to-end test 
that Director Kincannon spoke of this morning.
    Mr. Clay. Thank you for that. Mr. Romeo.
    Mr. Romeo. Thank you, Mr. Chairman. With the DADS system, 
because of the similarities to the receipt of data from the 
2000 system from the Census Bureau and that we will receive in 
the 2010 census, we're very confident that the testing will be 
adequate to ensure that the system is fully functional.
    Mr. Clay. What is the current earned value management data 
allowing regarding shelving, regarding constant schedule 
performance for the key acquisitions? Specifically, are you on 
schedule to deliver on your schedule estimate, Ms. Janey?
    Ms. Janey. As was discussed earlier, requirements have 
continued to evolve with the FDCA system and we are 
experiencing a--less than a 10 percent overrun on the project 
to date.
    Mr. Clay. Has the Bureau adequately defined specific 
requirements for the major system acquisitions that you are 
associated with? Has it been clear what they're purchasing?
    Ms. Janey. As Director Kincannon stated earlier, this is 
the first time a handheld has been used. It's involving a 
different part of the Bureau, in the field operations. So not 
surprisingly, there is some evolution to the requirements. That 
was not unexpected, but it's continued to go. What I'm pleased 
to tell you is that the Bureau is actively working to get to a 
point in the very near future where we lock down the 
requirements so that the requirements are set and we are then 
moving forward to the date that we are constantly reminded of.
    Mr. Clay. I know that the Harris Corp. has had several 
government contracts, I guess, in its history. And this is 
taxpayers' money. So I mean, it's not open-ended. And it ought 
to be guarded and we should be good stewards of it, all of us.
    Ms. Janey. Absolutely, Chairman Clay. We have--75 percent 
of Harris money goes to the government in one shape or form.
    Mr. Clay. All right. How about Ms. Marks, your earned value 
management data, what is it showing?
    Ms. Marks. Our earned value management data shows us on 
cost and on schedule within the life cycle budget for the DRIS 
program, and the Census Bureau has completed all requirements 
definition and they are firm.
    Mr. Clay. And that goes back to your history with the 
Bureau in, I guess, the 2000 census?
    Ms. Marks. I think the fact that we have personnel who have 
worked together, the fact that we have people who are skilled 
in the census domain practice and the fact that those are the 
people assigned to this project along with a wonderful team of 
subcontractors who are small and large businesses who all 
participated in the 2000 census helps us reduce risk and stay 
on schedule.
    Mr. Clay. Thank you for that response. Mr. Romeo.
    Mr. Romeo. We also are on budget and on schedule.
    Mr. Clay. Thank you for that response.
    If anyone else has anything to add?
    Ms. Marks. No, sir.
    Mr. Clay. If not, let me thank you, thank the panel for 
their indulgence today. And that ends the testimony of this 
panel. And without objection, the committee is adjourned. Thank 
you.
    [Whereupon, at 3:54 p.m., the subcommittee was adjourned.]

                                 
