2010 Census
Plans for Decennial Census Operations and Technology Have Progressed, But Much Uncertainty Remains Gao ID: GAO-08-886T June 11, 2008On April 3, 2008, the Secretary of Commerce announced significant changes to how the Census Bureau (Bureau) would conduct nonresponse follow-up, its largest field operation, in which census workers interview households that do not return initial census forms for the 2010 decennial census, and to its Field Data Collection Automation (FDCA) contract. The Bureau has since issued a redesigned plan to conduct a paper-based follow-up operation, an integrated 2010 Census project schedule, and is working on revising the FDCA contract. These are major changes late in the decennial census cycle. This testimony discusses (1) the Bureau's plans for conducting a paper-based nonresponse follow-up operation, (2) management of the FDCA contract and its latest cost estimates, and (3) the status of the Bureau's integrated 2010 project schedule. This testimony is based on past work, recent interviews with Bureau officials, and a review of redesign documents.
The Bureau has taken important steps to plan for a paper-based nonresponse follow-up operation, but several aspects remain uncertain. On May 8, 2008, the Bureau issued a paper-based nonresponse follow-up plan that details key components of the operation and describes processes for managing it and other operations. However, the plan envisions using an information system to manage the field operation workload, which experienced significant problems when tested earlier in the dress rehearsal. These problems make it more critical to test the system's capabilities for supporting the nonresponse follow-up operation. The Bureau will also institute new strategies--through second mailings and a new approach to remove late mail returns--but has only tested some aspects of these operations and will be unable to test them in a dress rehearsal, making it difficult to estimate their impact on operations in 2010. Ideally, the dress rehearsal should test almost all of the operations and procedures planned for the decennial under as close to census-like conditions as possible. Bureau officials expect that some small-scale testing will occur, particularly integration testing for its operations control system and cognitive testing of the forms used by enumerators for nonresponse follow-up, but what will be tested and when is not yet certain. The Bureau has taken several positive steps to address FDCA program management and oversight, but cost estimates need reconciling. The Bureau has taken actions to strengthen the FDCA program office leadership and expertise. To lead the program office, the Bureau has assigned an experienced Census program manager and hired an outside information technology expert to provide executive level guidance. The Bureau has also taken actions to improve communications and transparency of contractor activities. Further, the Bureau has obtained an independent government cost estimate based on the changes to the FDCA program's scope, which is nearly $600 million less than the contractor's rough order of magnitude estimate. After the contractor develops its detailed cost estimate, then the Bureau will need to reconcile the two cost estimates and renegotiate the contract. The Bureau will need to ensure that the final contract modifications and terms allow for FDCA program activities to be conducted in a timely and accurate manner for the 2010 decennial census. The Bureau's integrated schedule, dated May 22, 2008, identifies over 11,000 activities and milestones for the census. There is overlap in the testing and deployment schedule for the handheld device that will be used to collect address data in the field. Further, the Bureau's summary of key milestones does not include a milestone for when testing of key activities related to nonresponse follow-up will take place. Such milestones are important because nonresponse follow-up is the single largest field operation and will not be part of a dress rehearsal. The Bureau recognizes that it could include a key milestone for nonresponse follow-up testing activities. GAO is reviewing in greater detail the summary and integrated schedule of milestones and a summary of program risks provided on June 4th.
GAO-08-886T, 2010 Census: Plans for Decennial Census Operations and Technology Have Progressed, But Much Uncertainty Remains
This is the accessible text file for GAO report number GAO-08-886T
entitled '2010 Census: Plans for Decennial Census Operations and
Technology Have Progressed, But Much Uncertainty Remains' which was
released on June 12, 2008.
This text file was formatted by the U.S. Government Accountability
Office (GAO) to be accessible to users with visual impairments, as part
of a longer term project to improve GAO products' accessibility. Every
attempt has been made to maintain the structural and data integrity of
the original printed product. Accessibility features, such as text
descriptions of tables, consecutively numbered footnotes placed at the
end of the file, and the text of agency comment letters, are provided
but may not exactly duplicate the presentation or format of the printed
version. The portable document format (PDF) file is an exact electronic
replica of the printed version. We welcome your feedback. Please E-mail
your comments regarding the contents or accessibility features of this
document to Webmaster@gao.gov.
This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed
in its entirety without further permission from GAO. Because this work
may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this
material separately.
Testimony before the Committee on Oversight and Government Reform and
the Subcommittee on Information Policy, Census, and National Archives,
House of Representatives:
United States Government Accountability Office:
GAO:
For Release on Delivery:
Expected at 10:00 a.m. EDT:
Wednesday, June 11, 2008:
2010 Census:
Plans for Decennial Census Operations and Technology Have Progressed,
But Much Uncertainty Remains:
Statement of Mathew J. Scirč:
Director, Strategic Issues:
David A. Powner:
Director, Information Technology Management Issues:
GAO-08-886T:
GAO Highlights:
Highlights of GAO-08-886T, a testimony to the Committee on Oversight
and Government Reform and the Subcommittee on Information Policy,
Census, and National Archives, House of Representatives.
Why GAO Did This Study:
On April 3, 2008, the Secretary of Commerce announced significant
changes to how the Census Bureau (Bureau) would conduct nonresponse
follow-up, its largest field operation, in which census workers
interview households that do not return initial census forms for the
2010 decennial census, and to its Field Data Collection Automation
(FDCA) contract. The Bureau has since issued a redesigned plan to
conduct a paper-based follow-up operation, an integrated 2010 Census
project schedule, and is working on revising the FDCA contract. These
are major changes late in the decennial census cycle. This testimony
discusses (1) the Bureau‘s plans for conducting a paper-based
nonresponse follow-up operation, (2) management of the FDCA contract
and its latest cost estimates, and (3) the status of the Bureau‘s
integrated 2010 project schedule. This testimony is based on past work,
recent interviews with Bureau officials, and a review of redesign
documents.
What GAO Found:
The Bureau has taken important steps to plan for a paper-based
nonresponse follow-up operation, but several aspects remain uncertain.
On May 8, 2008, the Bureau issued a paper-based nonresponse follow-up
plan that details key components of the operation and describes
processes for managing it and other operations. However, the plan
envisions using an information system to manage the field operation
workload, which experienced significant problems when tested earlier in
the dress rehearsal. These problems make it more critical to test the
system‘s capabilities for supporting the nonresponse follow-up
operation. The Bureau will also institute new strategies”through second
mailings and a new approach to remove late mail returns”but has only
tested some aspects of these operations and will be unable to test them
in a dress rehearsal, making it difficult to estimate their impact on
operations in 2010. Ideally, the dress rehearsal should test almost all
of the operations and procedures planned for the decennial under as
close to census-like conditions as possible. Bureau officials expect
that some small-scale testing will occur, particularly integration
testing for its operations control system and cognitive testing of the
forms used by enumerators for nonresponse follow-up, but what will be
tested and when is not yet certain.
The Bureau has taken several positive steps to address FDCA program
management and oversight, but cost estimates need reconciling. The
Bureau has taken actions to strengthen the FDCA program office
leadership and expertise. To lead the program office, the Bureau has
assigned an experienced Census program manager and hired an outside
information technology expert to provide executive level guidance. The
Bureau has also taken actions to improve communications and
transparency of contractor activities. Further, the Bureau has obtained
an independent government cost estimate based on the changes to the
FDCA program‘s scope, which is nearly $600 million less than the
contractor‘s rough order of magnitude estimate. After the contractor
develops its detailed cost estimate, then the Bureau will need to
reconcile the two cost estimates and renegotiate the contract. The
Bureau will need to ensure that the final contract modifications and
terms allow for FDCA program activities to be conducted in a timely and
accurate manner for the 2010 decennial census.
The Bureau‘s integrated schedule, dated May 22, 2008, identifies over
11,000 activities and milestones for the census. There is overlap in
the testing and deployment schedule for the handheld device that will
be used to collect address data in the field. Further, the Bureau‘s
summary of key milestones does not include a milestone for when testing
of key activities related to nonresponse follow-up will take place.
Such milestones are important because nonresponse follow-up is the
single largest field operation and will not be part of a dress
rehearsal. The Bureau recognizes that it could include a key milestone
for nonresponse follow-up testing activities. GAO is reviewing in
greater detail the summary and integrated schedule of milestones and a
summary of program risks provided on June 4th.
What GAO Recommends:
At this time GAO is not making any new recommendations.
To view the full product, including the scope and methodology, click on
[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-08-886T]. For more
information, contact Mathew Scirč at (202) 512-6806 or sciremj@gao.gov
or David Powner at (202) 512-9286 or pownerd@gao.gov.
[End of section]
Mr. Chairmen and Members of the Committee and Subcommittee:
We are pleased to be here today to discuss the Census Bureau's (Bureau)
plans for conducting the decennial census. For 2010, the Bureau
intended to automate field data collection activities as a way to
reduce costs and improve data quality and operational efficiency;
however, testing uncovered several problems with its planned use of
technology and the Bureau has now revised its plans. This statement
focuses on the Bureau's efforts to redesign the 2010 Census, including
1) the Bureau's plans for conducting a paper-based nonresponse follow-
up operation in 2010; (2) management of the Field Data Collection
Automation (FDCA) program and its latest cost estimates; and (3) the
status of the Bureau's integrated 2010 schedule, including milestones.
In March 2008, we designated the 2010 Census as a high-risk area,
citing several long-standing and emerging challenges.[Footnote 1] These
challenges include weaknesses in managing information technology (IT),
questions surrounding the performance of handheld computers,
uncertainty over the cost of the 2010 Census, and the elimination of
several operations from the 2008 Dress Rehearsal. In February 2008, the
Director of the Bureau initiated a replanning of the FDCA program, a
major acquisition that includes systems, equipment (including handheld
computers), and infrastructure for field staff to use in collecting
data for the 2010 Census. After analyzing several options to revise the
design of the decennial, the Secretary of Commerce, on April 3, 2008,
announced that the Bureau would no longer use handheld computers in its
largest field operation, nonresponse follow-up--in which field workers
interview households that did not return census forms. However, the
Bureau would continue with the contract for the FDCA program to provide
handheld computers for address canvassing--in which field workers
verify addresses--and develop the information system for controlling
the workload of all census field operations. The Bureau estimated that,
along with updating its assumptions, the option of conducting a paper-
based nonresponse follow-up but using handheld computers for address
canvassing, would result in a cost increase of $2.2 billion to $3
billion over the previously reported estimate of $11.5 billion.
On April 15, 2008, the Bureau Director reported on the Bureau's ongoing
efforts to address problems associated with the FDCA program and its
plans to implement a paper-based nonresponse follow-up. In addition to
announcing strengthened management planning and oversight, he
reiterated that--from April 9, 2008 when the Director testified before
this Committee--the Bureau would provide a detailed operating plan for
its FDCA program within 30 days, including deadlines for key
milestones, and the related paper-based nonresponse follow-up
operation; in 45 days, the Bureau would develop an integrated project
schedule for the 2010 Census; and, in 60 days, the Bureau would produce
a testing program for the automated address canvassing operation.
Our testimony today is based on our past work, including our
observation of the use of handheld computers in the address canvassing
dress rehearsal, as well as the status of the Bureau's redesign
efforts. In assessing the status of the redesign, we reviewed and
discussed with Bureau officials documents, including plans for a paper-
based nonresponse follow-up operation, related FDCA documents, and the
2010 Census integrated schedule. This work was conducted in accordance
with generally accepted government auditing standards. Those standards
require that we plan and perform the audit to obtain sufficient,
appropriate evidence to provide a reasonable basis for our findings and
conclusions based on our audit objectives. We believe that the evidence
obtained provides a reasonable basis for our findings and conclusions.
In summary, the Bureau has taken important steps to plan for a paper-
based nonresponse follow-up operation, but several aspects remain
uncertain. The Bureau's plan for nonresponse follow-up, released on May
8, 2008, details key components of the operation and its management.
However, the plan envisions using an information system, to manage the
field operation workload, which experienced significant problems when
tested earlier in the dress rehearsal, and proposes new replacement
mailing and late mail return strategies, which have not been fully
tested. The Bureau has also taken several positive steps to address
FDCA program management and oversight, but will need to reconcile the
cost estimates from its own FDCA contractor and an independent
government estimate, which differed by nearly $600 million. Finally,
the Bureau's integrated schedule, dated May 22, 2008, identifies over
11,000 activities and milestones for the decennial. However, there is
overlap in the testing and deployment schedule for the handheld device
that will be used to collect address data in the field, and the
integrated schedule also does not specifically define testing for key
information technology systems (e.g. system, integration, and end-to-
end). The Bureau also issued the 2010 Census Key Operational Milestone
Schedule. This represents a higher level summary of about 175 key
activities and is linked to the more exhaustive integrated schedule.
However, there are several notable exceptions to this schedule of key
operational milestones. For example, the schedule does not include a
milestone for when testing of key activities related to nonresponse
follow-up will take place. Such milestones are important because
nonresponse follow-up is the single largest field operation and will
not be part of a dress rehearsal. The Bureau recognizes that it could
include a key milestone for nonresponse follow-up testing activities.
We are currently reviewing in greater detail the summary and integrated
schedule of milestones and the recently revised summary of program
risks provided on June 4th.
Background:
The Bureau has less than two years until Census Day. To ensure a
successful census, sound risk management will be crucial, particularly
given its scope, magnitude, and immutable deadlines of the census. The
size of the decennial operation means that small problems can magnify
quickly, and big problems could be overwhelming. For example, 60
seconds might seem like an inconsequential amount of time, but in 2000,
if enumerators had spent just 1 minute more at each household during
nonresponse follow-up, almost $10 million would have been added to the
cost of the census. Further, sound risk management is important to a
successful census because many risks are interrelated, and a
shortcoming in one operation could cause other operations to spiral
downward. For instance, a low mail response rate would drive up the
follow-up workload, which in turn would increase staffing needs and
costs. Of course, the reverse is also true, where a success in one
operation could positively affect downstream operations. Nevertheless,
rigorous up-front planning and testing, as well as risk mitigation
plans, are the best ways to stave off problems. Finally, the census is
conducted against a backdrop of immutable deadlines; the census'
elaborate chain of interrelated pre-and post-Census Day activities is
predicated upon those dates. To meet legally mandated reporting
requirements, including delivery of population counts to the President
on December 31, 2010, census activities need to take place at specific
times and in the proper sequence.
Bureau Has Taken Important Steps in Planning for a Paper-based
Nonresponse Follow-up Operation, But Much Remains Uncertain:
On May 8, 2008 the Bureau issued its plans for conducting the 2010
Census paper-based nonresponse follow-up operation outlining key
operational decisions. Among these is the need to develop an
information system to manage the workload for a paper-based nonresponse
follow-up operation and for additional field infrastructure, such as
more telephones and computers to support this operation, to restructure
the replacement mailing[Footnote 2] and the removal of late mail
returns from the nonresponse follow-up workload, as well as the need
for cognitive testing of the enumerator questionnaire used to collect
data from nonrespondents.
The contractor carrying out the FDCA program will develop the
operations control system, which is designed to manage field operations
that rely on paper as well as those that rely upon the handheld
computers. The Bureau is particularly concerned about this system
because when it was tested as part of earlier dress rehearsal
operations--for example, during group quarters validation--it was found
to be unreliable. As a result, the workload for these operations had to
be supplemented with additional paper-based efforts by local census
office staff, instead of electronically as intended. The operations
control system is critical because it is intended to provide managers
with essential real-time information such as enumerator productivity
and the status of workload such as interviews conducted and remaining.
Bureau officials said that the manual workaround was manageable for the
dress rehearsal with just two local census offices; however, such a
manual workaround would be nearly impossible to do when operations are
carried out nationwide next year. Officials said that they expect to
review computer screen shots of the operations control system reports
it will use to manage the nonresponse follow-up operation in January
2009; however, the Bureau has not yet determined when and how testing
of the operations control system before nonresponse follow-up, which
begins in April 2010, will occur.
The Bureau will be using newly developed systems for integrating
responses and managing nonresponse follow-up workload that have not yet
been fully tested in a census-like environment. The Bureau's contract
for the Decennial Response Integration System, designed to help
identify households that have not yet returned census forms and to
collect the results from enumerators conducting nonresponse follow-up
interviews, will process each mail return and enumerator questionnaire
and transmit to the FDCA program the number of questionnaires received.
In turn, FDCA will manage the nonresponse follow-up workload, in part
by removing initial late mail returns from the list of housing units
requiring follow-up visits. Consequently, depending on time and cost
considerations, Bureau officials believe that the Bureau must conduct,
at a minimum, a small scale simulation of the integration and
communication between the Decennial Response Integration System and
FDCA for such aspects as load testing for a paper-based operation, and
interfaces such as when the paper is processed by the Decennial
Response Integration System and when the check-in status is transmitted
to individual local census offices through management reports processed
by the FDCA program. When or how these tests will be completed is not
clear.
The Bureau's plans for nonresponse follow-up will also require changes
in local census office infrastructure. The Bureau expects it will need
additional hardware, including printing and scanning equipment,
computers, and telephones. Further, the Bureau expects to scale the
FDCA network to support a system for keying in large volumes of data
related to hiring and payroll for over 700,000 field workers it plans
to hire for the nonresponse follow-up operation. Previously, the Bureau
expected to maintain field worker time reporting using the handheld
computer. Also, the Bureau expected to hire fewer field workers.
The Bureau's redesign has also changed the replacement mailing strategy
which will be used in 2010. The replacement mailing is a second mailing
sent to nonresponding households. Testing has shown that a second
mailing increases the overall response rate and reduces costs by
increasing the number of returns that come in by mail, decreasing the
need for census field workers to collect census data in person. Prior
to the redesign, the Bureau planned to send second mailings to all
nonresponding households that initially received the census form in the
mail. However these plans changed, in part because, according to the
Bureau, without using handheld computers for nonresponse follow-up, it
would not be able to dynamically remove late mail returns--including
those resulting from the replacement mailing--from the enumerator
assignments on a daily basis. The Bureau had to devise a way to balance
the time available to print replacement questionnaires with the time
available to remove late mail returns from the paper-based nonresponse
follow-up workload.
The Bureau now plans a multi-part approach. First, it will send
approximately 25-30 million blanket replacement mailings to census
tracts with low response rates, based on historical response rate data
from 1990 and 2000 Census and the American Community Survey. As a
result, all housing units in these selected census tracts would receive
a second census form, regardless of whether or not they returned the
initial form. Similarly, the Bureau plans to target a second mailing to
an additional 15 million households in census tracts that are in the
middle-range of mail response rates. Finally, the Bureau will not send
a replacement mailing to households located in census tracts that
previously had high mail response rates. This combination "blanket" and
"targeted" mailing strategy is a new approach that will not be tested
prior to the 2010 Census. If the replacement mailing does not function
as planned, this strategy could confuse respondents in the blanket
mailing areas and result in multiple responses from the same household
that return both forms. It is instructive to consider that the Bureau's
previous experience with a blanket second "replacement" questionnaire
sent to all housing units located in the 1998 dress rehearsal sites
caused a significant number of households with multiple responses. As a
result, the replacement mailing was dropped from the 2000 Census design
because the Bureau was concerned that it would have been overwhelming
to process multiple census responses during the actual census.
Moreover, without the benefit of implementing nonresponse follow-up
during the dress rehearsal, the Bureau will not know how well its new
system for removal of late mail returns will work. While the Bureau
encourages respondents to mail back their census forms quickly, some
are not returned until the middle of April or later, after the
nonresponse follow-up operation has begun. To reduce the cost of
nonresponse follow-up and to minimize respondent burden, it is
beneficial to the Bureau to remove these late mail returns from the
nonresponse follow-up universe. Because nonresponse follow-up will be
paper-based rather than conducted with handheld computers, the Bureau
will remove late mail returns with the FDCA program prior to April 20
and manually thereafter; however, the recent Bureau plans provide only
timelines for removing late mail returns and the Bureau has not yet
finalized the workload estimates or how it will manage this work. Not
having an opportunity to rehearse its strategy for removing late mail
returns makes difficult any estimate of resulting workload.
In addition, Bureau officials said that it will be important to conduct
cognitive testing of the questionnaire used by enumerators for
nonresponse follow-up. With the change from using handheld computers, a
paper questionnaire will be used by census enumerators in the 2010
nonresponse follow-up when making personal visits to housing units to
collect census data. When developing this questionnaire, the Bureau
plans to draw upon its extensive research and testing of interviewer-
conducted questionnaires developed for other censuses and surveys as
well as lessons learned in Census 2000. According to its May 8, 2008
plans for conducting the paper-based nonresponse follow-up, the Bureau
will conduct this cognitive and usability testing in early summer 2008
and the testing will address both respondent interactions and ease of
use for the census enumerators. The Bureau expects the questionnaire
will have space for up to six people as in Census 2000 and will link
other household members to the address via a continuation form; include
coverage questions; meet the Decennial Response Integration System data
capture specifications; and collect data on the outcome of the
enumeration.
Not being able to test the paper-based nonresponse follow-up in the
2008 Dress Rehearsal introduces risk because the dress rehearsal will
no longer be a dry-run of the decennial census. While the Bureau has
carried out a paper-based follow-up operation in the past, there are
now new procedures and system interfaces that, as a result of its
exclusion from the dress rehearsal, will not be tested under census-
like conditions. We discussed the nonresponse follow-up plan with
Bureau officials and they acknowledge the importance of testing new and
changed activities of nonresponse follow-up as well as system
interfaces to reduce risk. However, because plans have changed for many
aspects of the nonresponse follow-up operation, Bureau officials are
uncertain about testing and are still trying to determine which
activities and interfaces will be tested and when that testing will
occur.
It is important to note that the Bureau has taken some important
initial steps to manage the replannning effort. For example, the Bureau
has added temporary "action officers" to its 2010 governance structure.
As of April 17, 2008, six action officers had been identified to
achieve the six objectives in its Recovery Plan--nonresponse follow-up
replan, reduce FDCA risk, improve communications, document decennial
program testing, improve program management, and baseline an integrated
schedule. Each action officer is assigned to one of the objectives.
These action officers are intended to be catalysts, liaisons, and
facilitators responsible for ensuring that the tasks and milestones for
each objective are met. Also, the action officers meet with the
Associate and Assistant Directors to facilitate quick decision-making
and on a regular basis provide updates on the status of plans. Weekly,
the Bureau's Director meets with the Department of Commerce's Deputy
Secretary to discuss the status of the replan for the 2010 Census.
The Bureau has also issued documents that describe actions it will take
to identify and manage risk. The Bureau's 2010 Census Program
Management Plan, issued May 5, 2008, contains information about the
risk management process and notes that 24 program-level or high level-
risks have been identified, were currently being validated, and that
each of these 24 risks would have either mitigation or contingency
plans associated with them. However, according to Bureau officials,
these 24 risks were associated with an automated operation and the
Bureau had not yet developed risks related to the paper-based
nonresponse follow-up operation. We requested information on these 24
risks, and on June 4, 2008, the Bureau provided us with an updated
program-level risk document. The update now includes 25 program-level
risks and identifies several risks related to the redesign including
late design changes and testing. However, the Bureau has not updated
project-level risks--which are risks specific to an operation or
system--for nonresponse follow-up since the change to paper was
announced. Once the Bureau provides project-level risk documents, we
will assess the Bureau's actions to identify, prioritize, and manage
risk for the replanned nonresponse follow-up operation.
Bureau Has Improved Program Management and Oversight, but Cost
Estimates Need Timely Reconciling:
The Bureau has taken steps to strengthen the FDCA program office
leadership and expertise. The Bureau has recently assigned an
experienced Bureau manager to manage the FDCA program office. According
to the Bureau, the manager has extensive experience in directing major
IT projects. The Bureau has also hired an outside IT expert, to provide
advice and guidance to the FDCA program office. The Bureau has also
implemented key activities to help improve management and transparency
of contractor activities. Bureau officials have established a schedule
for daily assessment meetings with contractor personnel; are conducting
weekly status assessment and resolution meetings with the Deputy
Director and Director; and are holding regular meetings with the
Department of Commerce.
The Bureau has obtained cost estimates for FDCA from both Harris and
MITRE, based on the recent changes to the scope of the program. In
particular, these cost estimates include the January 16, 2008
requirements and the decision for a paper-based nonresponse follow-up
operation. Harris is estimating that the revised FDCA program will cost
roughly $1.3 billion; however, this cost estimate is preliminary and
expected to be further refined.
At the direction of the Bureau, MITRE developed an independent
government cost estimate in April 2008. MITRE's estimate is about $726
million, which is nearly $600 million less than the contractor's rough
order of magnitude estimate. A comparison of the two estimates reveals
significant differences in two areas: software development and common
support. In particular, Harris is estimating that software development
will be about $200 million greater than MITRE's independent estimate;
and that common support will be about $300 million greater than MITRE's
estimate.
* Software development ($200 million difference): MITRE officials noted
that these differences could be attributed to different assumptions
based on abnormal software development (such as starts and stops due to
budget instability), labor rates used, amount of additional staff
needed in order to maintain the schedule and to address quality and
testing issues, as well as cost contingency reserves.
* Common support ($300 million difference): Although this program
element contains the largest cost difference, MITRE officials noted
that they could not identify the primary cost drivers that caused the
gap. However, possible explanations could be cost contingency reserves
that may have been built into the Harris estimate, labor rates used,
unexpected high level of change management personnel resulting from
budget and requirements changes, and other potential impacts on
management resulting from program instability.
Harris had originally planned to deliver the cost estimate by August
20, 2008. However, the Bureau requested that this estimate be delivered
sooner and Harris recently agreed to deliver this cost estimate by July
15, 2008. The Bureau and contractor plan to reconcile and agree to a
final estimate by August 15, 2008. We plan to analyze the independent
cost estimate and the Harris final estimate for the program. As part of
this analysis, we intend to evaluate the methodology, as well as
underlying assumptions, used to develop each estimate.
The Bureau needs to act swiftly to finalize the FDCA program's cost
estimate and renegotiate the contract. In particular, it will need to
have a final cost estimate from Harris in mid-July, and will need to
reconcile this estimate with MITRE's independent estimate thoroughly
and quickly to have a final cost estimate by August 15, 2008. Our body
of work on the lessons learned on other major IT acquisitions,
highlights the importance of establishing realistic cost estimates
(through reconciliation of program and independent cost estimates),
using fixed price contract techniques for low risk procurement areas,
where appropriate, and establishing management reserve funds for
unexpected costs. In moving forward, it is important that the Bureau
exercise diligence in finalizing the contract terms to ensure that the
FDCA program is conducted in a timely and efficient manner for the 2010
decennial.
Bureau's Integrated Schedule Identified Activities and Associated
Milestones but Did Not Address Risks and Costs:
The Bureau designed its 2010 Census Integrated Schedule, dated May 22,
2008, to provide information on its schedule framework and activity-
level design as well as to describe the program complexity and methods
that the Bureau will use to manage the 44 interdependent operations,
incorporating over 11,000 unique activities, to conduct the 2010
Census. The Bureau briefed committee staff and us on this final
integrated schedule last week. Based on this briefing and our
preliminary review of the schedule, we can offer some observations.
The integrated schedule does identify activities that need to be
accomplished for the decennial and the Bureau establishes milestones
for completing tasks. However, the schedule does not link those
activities with associated risks nor does it capture the cost of
operations. We previously recommended to manage the 2010 Census and
contain costs, the Bureau develop a comprehensive, integrated project
plan for the 2010 Census that should include risk and mitigation plans,
updated cost estimates, and detailed milestones that identify all
significant relationships.[Footnote 3] We also observed that testing
the handheld computer that will be used in the address canvassing
operation--an activity we have previously identified as important in
mitigating risks associated with use of new technology--overlapped with
its deployment. Specifically, in describing the testing and integrating
of handheld computers for the address canvassing operation, the
schedule indicates that this activity will begin in December 2008 and
be completed in late March 2009; however, the deployment of the
handheld device for address canvassing will actually start in February
2009, before the completion of testing and integration. It would appear
uncertain that the testing and integration milestones would permit
modification to technology or operations prior to the onset of
operations. Further, the Bureau's integrated schedule does not
specifically define testing (e.g., system, integration, and end-to-
end). Separately, the Bureau on June 6, 200 produced a testing plan for
the address canvassing operation.
On May 22, 2008, the Bureau also issued the 2010 Census Key Operational
Milestone Schedule. This represents a higher level summary of key
operations and is linked to the more exhaustive integrated schedule.
The Bureau identified about 175 activities that it considers key and
that are used by senior management to oversee the 2010 Census. However,
there are several notable exceptions to this schedule of key
operational milestones. For example, there is no key milestone for
identification of program and project risks in light of the significant
change in planned operations, nor for developing necessary mitigation
or contingency plans. Including key milestones for risk identification
and mitigation in its high-level schedule will enable the Bureau to
stay focused on activities which can directly impact the quality or
cost of the 2010 Census. Nor does the schedule include a milestone for
when testing of key activities related to nonresponse follow-up will
take place. This is despite the fact that this represents the single
largest field operation and will not be part of a dress rehearsal. The
Bureau does recognize that it could include in its high-level summary
schedule a key milestone for nonresponse follow-up testing activities.
Further testing schedules for address canvassing and the operations
control system also do not appear as key milestones, though they do
appear in the detailed integrated schedule. Including these critical
activities as part of the list of key milestones could ensure greater
management attention, as well as help in focus oversight. We are
currently reviewing in greater detail the summary and integrated
schedule of milestones and the recently revised program-level risk
document provided on June 4, 2008.
In summary, the Bureau has taken some important steps toward managing
the changes it plans for conducting the 2010 Census. Yet much remains
uncertain and in the absence of a full dress rehearsal, the risks to a
successful decennial census are substantial. Risks are especially high
for the 2010 Census nonresponse follow-up operation both because the
Bureau will not reap the benefits of having a dress rehearsal for this
key operation but also because it is changing its approach late in the
decade. These make even more compelling the need for the Bureau to
specify what tests it plans to conduct in the absence of a dress
rehearsal and when such testing will take place.
The Bureau will also need to take several next steps to finalize the
FDCA program's cost estimate. In particular, it will need to have a
final cost estimate from Harris, as soon as possible, in order to have
a sufficient amount of time to complete modifications to the contract
by the end of the fiscal year. Our body of work on the lessons learned
on other major IT acquisitions, highlights the importance of
establishing realistic cost estimates (through reconciliation of
program and independent cost estimates), using fixed price contract
techniques for low risk procurement areas, such as hardware, and
establishing management reserve funds for unexpected costs. In moving
forward, it is important that the Bureau exercise diligence in
finalizing the contract terms to ensure that the FDCA program can be
conducted in a timely and efficient manner.
Finally, the Bureau has developed a detailed integrated schedule of
activities that need to be conducted during the 2010 Census and
established milestones for completing them. It will be important for
the Bureau to ensure that among the key milestones and activities that
are highlighted for management and oversight are those that represent
the greatest impact on the ultimate cost and quality of the 2010
Census.
Mr. Chairmen and members of the committee and subcommittee, this
concludes our statement. We would be happy to respond to any questions
that you or members of the subcommittee may have at this time.
If you have any questions on matters discussed in this testimony,
please contact Mathew J. Scirč at (202) 512-6806 or David A. Powner at
(202) 512-9286 or by email at sciremj@gao.gov or pownerd@gao.gov. Other
key contributors to this testimony include Carol Cha, Betty Clark,
Vijay D'Souza, Sarah Farkas, Richard Hung, Andrea Levine, Catherine
Myrick, Lisa Pearson, Cynthia Scott, and Niti Tandon.
[End of testimony]
Footnotes:
[1] GAO, Information Technology: Significant Problems of Critical
Automation Program Contribute to Risks Facing 2010 Census, [hyperlink,
http://www.gao.gov/cgi-bin/getrpt?GAO-08-550T] (Washington, D.C.:
Mar. 5, 2008). See also GAO, Census 2010: Census at Critical Juncture
for Implementing Risk Reduction Strategies, [hyperlink,
http://www.gao.gov/cgi-bin/getrpt?GAO-08-659T] (Washington, D.C.:
Apr. 9, 2008).
[2] A replacement mailing is a replacement questionnaire sent to
households to remind and encourage them to return their census
questionnaire.
[3] GAO, 2010 Census: Cost and Design Issues Need to Be Addressed Soon,
[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-04-37]
(Washington, D.C.: Jan. 15, 2004).
[End of section]
GAO's Mission:
The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people.
GAO examines the use of public funds; evaluates federal programs and
policies; and provides analyses, recommendations, and other assistance
to help Congress make informed oversight, policy, and funding
decisions. GAO's commitment to good government is reflected in its core
values of accountability, integrity, and reliability.
Obtaining Copies of GAO Reports and Testimony:
The fastest and easiest way to obtain copies of GAO documents at no
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each
weekday, GAO posts newly released reports, testimony, and
correspondence on its Web site. To have GAO e-mail you a list of newly
posted products every afternoon, go to [hyperlink, http://www.gao.gov]
and select "E-mail Updates."
Order by Mail or Phone:
The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or
more copies mailed to a single address are discounted 25 percent.
Orders should be sent to:
U.S. Government Accountability Office:
441 G Street NW, Room LM:
Washington, D.C. 20548:
To order by Phone:
Voice: (202) 512-6000:
TDD: (202) 512-2537:
Fax: (202) 512-6061:
To Report Fraud, Waste, and Abuse in Federal Programs:
Contact:
Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]:
E-mail: fraudnet@gao.gov:
Automated answering system: (800) 424-5454 or (202) 512-7470:
Congressional Relations:
Ralph Dawn, Managing Director, dawnr@gao.gov:
(202) 512-4400:
U.S. Government Accountability Office:
441 G Street NW, Room 7125:
Washington, D.C. 20548:
Public Affairs:
Chuck Young, Managing Director, youngc1@gao.gov:
(202) 512-4800:
U.S. Government Accountability Office:
441 G Street NW, Room 7149:
Washington, D.C. 20548:
