[Senate Hearing 114-436]
[From the U.S. Government Publishing Office]
S. Hrg. 114-436
2020 CENSUS: CHALLENGES FACING THE BUREAU FOR A MODERN, COST EFFECTIVE
SURVEY
=======================================================================
HEARING
before the
COMMITTEE ON
HOMELAND SECURITY AND GOVERNMENTAL AFFAIRS
UNITED STATES SENATE
ONE HUNDRED FOURTEENTH CONGRESS
FIRST SESSION
__________
APRIL 20, 2015
__________
Available via the World Wide Web: http://www.fdsys.gov/
Printed for the use of the
Committee on Homeland Security and Governmental Affairs
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
U.S. GOVERNMENT PUBLISHING OFFICE
94-905 PDF WASHINGTON : 2016
-----------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Publishing
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 HOMELAND SECURITY AND GOVERNMENTAL AFFAIRS
RON JOHNSON, Wisconsin Chairman
JOHN McCAIN, Arizona THOMAS R. CARPER, Delaware
ROB PORTMAN, Ohio CLAIRE McCASKILL, Missouri
RAND PAUL, Kentucky JON TESTER, Montana
JAMES LANKFORD, Oklahoma TAMMY BALDWIN, Wisconsin
MICHAEL B. ENZI, Wyoming HEIDI HEITKAMP, North Dakota
KELLY AYOTTE, New Hampshire CORY A. BOOKER, New Jersey
JONI ERNST, Iowa GARY C. PETERS, Michigan
BEN SASSE, Nebraska
Keith B. Ashdown, Staff Director
Courtney J. Allen, Counsel
Gabrielle A. Batkin. Minority Staff Director
John P. Kilvington, Minority Deputy Staff Director
Troy H. Cribb, Minority Chief Counsel for Governmental Affairs
Deirdre G. Armstrong, Minority Professional Staff Member
Laura W. Kilbride, Chief Clerk
Lauren M. Corcoran, Hearing Clerk
C O N T E N T S
------
Opening statements:
Page
Senator Lankford............................................. 1
Senator Carper............................................... 10
Senator Ayotte............................................... 12
Senator Peters............................................... 15
Prepared statements:
Senator Johnson.............................................. 33
Senator Carper............................................... 35
Senator Lankford............................................. 37
WITNESSES
Monday, April 20, 2015
Hon. John H. Thompson, Director, U.S. Census Bureau, U.S.
Department of Commerce......................................... 3
Robert Goldenkoff, Director of Strategic Issues, U.S. Government
Accountability Office.......................................... 6
Carol R. Cha, Director of Information Technology Acquisition
Management Issues, U.S. Government Accountability Office....... 8
Alphabetical List of Witnesses
Cha, Carol R.:
Testimony.................................................... 8
Joint prepared statement..................................... 45
Goldenkoff, Robert:
Testimony.................................................... 6
Joint prepared statement..................................... 45
Thompson, Hon. John H.:
Testimony.................................................... 3
Prepared statement........................................... 38
APPENDIX
Response to post-hearing questions submitted by Mr. Thompson..... 65
2020 CENSUS: CHALLENGES FACING THE
BUREAU FOR A MODERN, COST-EFFECTIVE SURVEY
----------
MONDAY, APRIL 20, 2015
U.S. Senate,
Committee on Homeland Security
and Governmental Affairs,
Washington, DC.
The Committee met, pursuant to notice, at 3:01 p.m., in
room SD-342, Dirksen Senate Office Building, Hon. James
Lankford, presiding.
Present: Senators Johnson, Lankford, Ayotte, Carper,
McCaskill, and Peters.
OPENING STATEMENT OF SENATOR LANKFORD
Senator Lankford. This hearing will come to order. I wanted
to do an opening statement, and then Senator Carper is on his
way. Senator Johnson is also on his way, and so as they slip
in, we will recognize them at the appropriate time as well. I
want to have this Committee come to order so we can begin this
process on time today.
I would like to welcome everyone to this afternoon's
hearing on the ``2020 Census: Challenges Facing the Bureau for
a Modern, Cost-Effective Survey.'' I will begin by recognizing
myself for this opening statement.
Director Thompson, let me first say to you on behalf of the
Committee, we would like to express all of our condolences on
the loss of Lawrence Buckner. We lost him in the line of duty
on April 9 in Census Bureau headquarters in Suitland, Maryland.
Know that our thoughts and prayers are with you and the agency
as you all walk through this time together. It was a senseless
act of violence, and we will all walk through it together with
his family.
The United States Constitution mandates that an actual
enumeration be made within 3 years after the first meeting of
the Congress of the United States and every subsequent 10
years.
As the 2020 decennial census approaches, I appreciate the
fact that the Census Bureau is actually taking steps to prepare
since, as we have seen in the past, a failure to do so can
drastically increase costs for the Federal Government and
ultimately the taxpayers. It is laudable that the Bureau is
working toward implementation of an innovation strategy with
the goal of saving the taxpayers' money.
However, as the Government Accountability Office (GAO), has
noted, although progress has been made since 2010, more work is
needed to ensure that technology systems are fully operational
and information remains secure and confidential.
While it is important that we have information regarding
the makeup of this Nation, it is also important that we
safeguard the privacy of the American people. Questions asked
of citizens of this Nation must be cost-effective and should
not be overly intrusive. When the questions go too far and
exceed what we should ask from the U.S. Government, it is
invasive and fuels an environment of distrust.
It is important that we examine not just the preparation
for the 2020 Census itself, but also the manner in which the
Census and other surveys, like the American Community Survey
(ACS), are being conducted, as well the substantive nature of
the questions asked. I look forward to the testimony, and I
will acknowledge the Ranking Member Senator Carper for his
opening statement when he arrives as well.
Let me introduce the witnesses. Then I am going to swear
the witnesses in as we go from here.
John Thompson is the Director of the Census Bureau. Before
his appointment as Director, Mr. Thompson was the President and
Chief Executive Officer (CEO) of the National Opinion Research
Center (NORC). Thanks again for being here.
Robert Goldenkoff is the Director of Strategic Issues at
the GAO. Mr. Goldenkoff leads reviews of the governmentwide
civil service reforms and ways of improving the cost-
effectiveness of the Federal statistical system.
Carol Cha is the Director of Information Technology
Acquisition Management Issues at the Government Accountability
Office. Before joining GAO, Ms. Cha led numerous reviews of
information technology (IT) systems at Federal agencies,
including the Departments of Commerce, Defense (DOD), and
Homeland Security (DHS).
Thank you, all three of you, for being here, and thanks for
your testimony, both your written testimony and the oral
testimony that you are about to begin as well.
I would like to ask unanimous consent that Senator
Johnson's statement be included in the record.\1\ Without
objection, so ordered.
---------------------------------------------------------------------------
\1\ The prepared statement of Senator Johnson appears in the
Appendix on page 33.
---------------------------------------------------------------------------
It is the tradition of this Committee that we swear in all
witnesses, so would you please rise and raise your right hand?
Do you swear that the testimony you will give before this
Committee will be the truth, the whole truth, and nothing but
the truth, so help you, God?
Mr. Thompson. I do.
Mr. Goldenkoff. I do.
Ms. Cha. I do.
Senator Lankford. Thank you. You may be seated.
Let the record reflect that the witnesses have answered in
the affirmative.
I will ask that each of you will give your testimony. You
have plenty of time to be able to do that. We are glad that you
came and you came well prepared on that. When Senator Carper
comes, I will recognize him for an opening statement. That may
be in between one of you, but I will not interrupt you in the
middle of your statement if he comes in during that time.
Mr. Thompson, you are first. Thank you. We would be glad to
receive your testimony now.
TESTIMONY OF HON. JOHN H. THOMPSON,\1\ DIRECTOR, U.S. CENSUS
BUREAU, U.S. DEPARTMENT OF COMMERCE
Mr. Thompson. Thank you, Senator Lankford. I would also
like to thank Chairman Johnson and Ranking Member Carper and
the Committee for allowing me this opportunity to testify on
our preparations for the 2020 Census.
---------------------------------------------------------------------------
\1\ The prepared statement of Mr. Thompson appears in the Appendix
on page 38.
---------------------------------------------------------------------------
Before I testify, however, I would like to take a moment to
honor a true hero, Officer Lawrence Buckner, who lost his life
protecting all of us at the U.S. Census Bureau. We are
profoundly saddened by the loss of Officer Buckner. No one can
ever measure the loss to his family. But our hearts are full,
and the Census Bureau extends its deepest sympathy to his wife,
Linda, his son, Lawrence, and his family. Thank you, and we
mourn this tragic loss.
I will now proceed with my testimony. The Census Bureau is
fundamentally changing the way we will conduct the decennial
census, which is the largest civilian mobilization in the
United States. These efforts began earlier this decade by
establishing a goal to design and conduct the 2020 Census to
cost less per housing unit than the 2010 Census, while
maintaining the highest levels of quality.
The Census Bureau then identified the major cost drivers,
and with the Congress' support, we began researching and
testing major innovations oriented around technology and the
strategic use of information to rein in these cost drivers.
Today I will describe our work in four key innovation areas
that will lead to the reengineered 2020 Census. We believe that
investing now in these four key innovation areas can yield up
to $5 billion in savings relative to repeating the 2010 Census
in 2020. The tests we conducted in 2013 and 2014 and the four
tests we are conducting this year are informing the 2020 Census
design decisions, which we will deliver later this year. I will
discuss how these tests informed our planning and how the next
2 years position us for the critical end-to-end test in 2018.
The alternative is repeating the 2010 Census, which would
forfeit the savings of $5 billion.
There are four key areas of innovation and potential
savings:
First, better address validation. By using the U.S. Postal
Service (USPS) and other information sources, including aerial
imagery, we plan to avoid walking every street in the Nation to
validate the address list. We can save $1 billion through these
efforts.
Second, better response options. By making responding to
the Census more convenient through the Internet, phone, or by
mail, we can potentially save approximately $550 million.
Third, better use of existing information. By using
existing government and commercial information to reduce the
need to follow-up with non-responding housing units, we can
potentially save $1.2 billion.
Fourth, better field operations. By using technology to
manage and track cases, as well as to route the Census takers
who will be using smartphones and tablets rather than pencil
and paper, we can potentially save $2.3 billion.
As I noted above, the total savings we expect from these
four areas of innovation is $5 billion. I will now discuss each
of these innovation areas in more detail.
The foundation of an accurate Census is an accurate address
list, which includes both the address and the geospatial
location. Over the past few years, there have been tremendous
technological advances in the geospatial field. More and more
data become available, meaning that we no longer have to
validate every address by a personal visit. We are now
examining how to refine our procedures to integrate private
sector data and services to update our geospatial assets.
Specifically, we want to purchase address, road, and satellite
imagery instead of physically walking the entire United States.
This year's Address Validation Test encompasses two
components that will also bring insights into how we can build
a better address frame. The first component is to assess the
ability of statistical models to predict change, such as new
roads, new housing units, or other changes. The second
component will provide measures of the current accuracy of our
address list and our geospatial database.
The second area of innovation is developing better response
options. We are moving away from relying solely on the mailed
questionnaire and enumerator to count every household. We are
expanding options for people to self-respond by the Internet,
at home or on a mobile device remotely, as well as by
telephone. However, it is important to note that paper will
continue to be an option.
We want to make the Census as mobile and convenient as
possible. This means allowing respondents to answer the Census
without entering a Census ID. In fact, we anticipate that by
promoting the Internet option, there will be a dramatic
increase in the number of these responses in 2020. For this to
work, we must validate these responses quickly in real time.
The third area of innovation is better use of existing
information. The increased use of administrative records from
other Federal and State government agencies and third-party
commercial data can reduce costs. The use of administrative
records is not new to the decennial census, and we want to
expand our use of these data for the 2020 Census. We are
exploring several options, and two of the most promising are
using those data to help manage and even reduce the field
workload.
The two most significant areas of innovation and cost
savings are removing vacant units and using existing
information on persons to enumerate occupied housing units,
thereby removing them from the non-response follow-up
operations.
For example, during the 2010 Census, the field workload
included 50 million housing units. Each housing unit received
at least one in-person visit. Of these, 19 million were either
vacant or no longer existed. By using administrative data from
the Postal Service and other agencies, we believe we can
identify these vacant and non-existent housing units and remove
them from the in-person follow-up workload, achieving
substantial cost savings.
Administrative records may also us to enumerate occupied
housing units rather than enumerate them directly, and I would
like to have a conversation with the Congress about this
potential step.
Ultimately, we will have to send Census takers into the
field to enumerate the remaining non-responding households,
which underscores the importance of the fourth area of
innovation: better field operations. As part of the research
and testing, including the 2015 Census test in Maricopa County,
Arizona, we are examining our field staffing structure and
testing several technological innovations. The goal of
reengineering our field operations is to use technology more
efficiently and effectively to conduct and manage the 2020
Census field workload.
In previous Censuses, the entire process, both data
collection and management, was conducted by paper and pencil.
To measure progress, we had to rely on daily in-person meetings
with field staff and had no ability for real-time
communication. This paper-based data collection process was a
significant contributor to the overall cost increases of the
previous Census field operations.
We are developing a sophisticated operational control
system that will manage tasks and assignments in real time. We
intend to send our interviewers out with mobile devices rather
than paper and pencil. They will use these devices to collect
responses and report their time and attendance instead of using
the paper forms as in 2010. And we will have real-time measures
of progress.
Our goals are to incorporate operational best practices,
including the optimization of daily assignments, intelligent
routing, and real-time issue management. We are working with
the private sector as we build these systems.
The four key innovation areas represent significant cost
savings that can only be achieved if we get the opportunity to
complete significant testing and development in the next 2
years. We have very little time left to test and ultimately
important these innovations before we reach 2018, when the cost
of the Census will rise. 2018 is also important because we must
conduct a complete end-to-end test so we can be confident that
when we go live in 2020, all of our systems will work.
To realize the most modern decennial census ever, we have
to stay on track. That is why next year is crucial to the 2020
Census for the development of the key systems to support the
infrastructure to handle data collection and processing.
One of the critical activities in 2016 is the continued
development of our Census Enterprise system that will support
not just the 2020 Census but all of our data collection
activities at the Census Bureau. We will no longer build a
system that we will throw away after the decennial, as we have
for each previous Census. Known as the Census Enterprise Data
Collection and Processing (CEDCaP) program, we are moving to a
smarter, more cost-efficient, enterprise-level strategy to
manage core information technology aspects for all of our
Censuses and surveys.
At the core of this modern system that we are building is
our continued commitment to protecting privacy and
confidentiality of individuals' information. Confidentiality
and privacy is the very core of the Census Bureau's mission and
deeply ingrained in our culture. We protect the information the
public provides with a robust, comprehensive, and layered
cybersecurity system. We are actively engaged with the National
Institute of Standards and Technology (NIST) and the Department
of Homeland Security on these efforts.
From the Census Enterprise system and cybersecurity to each
test, each activity plays a significant role helping the Census
Bureau design an accurate and cost-effective 2020 Census.
Scaling up to the decennial census is complex, ranging over
many years and many operations, which must be synchronized to
meet our ultimate mandate. The $5 billion in savings cannot be
achieved without rigorous testing designed to inform
development of the systems and operations. We need your support
to achieve these goals.
Thank you, and I hope this update has been informative, and
I look forward to answering your questions.
Senator Lankford. Mr. Goldenkoff.
TESTIMONY OF ROBERT GOLDENKOFF,\1\ DIRECTOR OF STRATEGIC
ISSUES, U.S. GOVERNMENT ACCOUNTABILITY OFFICE
Mr. Goldenkoff. Thank you, Senator Lankford. I would like
to thank you and Chairman Johnson and Ranking Member Carper for
the opportunity to be here this afternoon to discuss the Census
Bureau's progress in developing a more cost-effective approach
to counting the Nation's population.
---------------------------------------------------------------------------
\1\ The joint prepared statement of Mr. Goldenkoff and Ms. Cha
appears in the Appendix on page 45.
---------------------------------------------------------------------------
As you know, this month the Bureau marked the midway point
in the decade-long countdown to Census Day 2020, giving greater
urgency and importance to the testing, operational, and
procurement decisions that it will make in the months ahead.
For example, this September, the Bureau plans to announce its
preliminary design for the 2020 Census, and two key field tests
to inform that decision are currently underway.
The cost of the decennial census has continually increased
during the past 40 years, in part because the Nation's
population has grown steadily larger, more diverse, and
increasingly difficult to enumerate. At about $13 billion, the
2010 Census was the costliest U.S. Census in history and was 56
percent more expensive than the $8.1 billion spent on the 2000
Census in constant 2010 dollars.
In my remarks today, I will discuss the Bureau's progress
in implementing four critical cost-savings initiatives and
their associated challenges. The four areas include: using data
previously provided to the government to help enumerate the
population; new processes for updating the Bureau's address
list and maps; reengineering field operations; and maximizing
self-response.
As the Director said, combined, the Bureau estimates that
these efforts could generate up to $5 billion in cost savings
and enable the Bureau to conduct the 2020 Census at a total
life-cycle cost of around $12.7 billion, or about the same as
the 2020 head count.
With respect to using data previously provided to the
government, the Bureau estimates that this initiative,
sometimes referred to as ``using administrative records,''
could save as much as $1.2 billion by reducing the need for
costly, labor-intensive follow-up work with non-responding
households.
The Bureau is testing the extent to which it can use
Federal data such as Social Security and Medicare records as
well as records from State, local, and tribal governments and
commercial sources to reduce the number of in-person visits,
local Census offices, and operations needed to ensure a
complete count.
However, before it will be able to realize cost savings or
improvements in data reliability from the use of these records,
the Bureau will first need to address such questions as the
quality of the records and whether they will meet the Bureau's
needs. For example, while race and ethnicity data are collected
in the Census, certain records available to the Bureau do not
include this information.
To reduce the cost of its efforts to update its address
list and maps by as much as $1 billion, the Bureau is testing
whether it can rely mainly on data from government agencies at
all levels to share and continuously update their address list
and street data with the Bureau.
In prior decennials, thousands of Bureau field staff walked
almost every street in the Nation as one of several operations
to update the Bureau's address list. Key questions here include
which map and address data sources are the most cost-effective
and whether the Bureau can accurately target its address
canvassing efforts.
With respect to improving the management of its field
operations, the Bureau is examining, among other things, how
best to automate enumerators' work, which could save an
estimated $2.3 billion. However, the Bureau must first resolve
whether it can fully test all the systems and procedures in
time for 2020.
A fourth cost-saving area involves maximizing self-response
through enhanced outreach and an Internet response option. This
effort could reduce the need for enumerators to visit non-
responding households and save around $500 million. However,
among other issues, the Bureau has yet to establish reliable
estimates of how much it will cost to deliver an Internet
response option and does not have integrated schedules for
completing the work.
The Bureau has identified or acknowledged many of the
challenges and questions associated with these four initiatives
and is working to address them. Because of their interrelated
nature, shortcomings in any one area could impact the success
of the others and thus put estimated cost savings and the
accuracy of the count at risk.
Going forward, it will be important to ensure the research
and testing for these initiatives continue as planned to inform
key design decisions later this year.
This concludes my prepared remarks, and I will be happy to
answer any questions that you may have.
Senator Lankford. Thank you.
Ms. Cha.
TESTIMONY OF CAROL R. CHA,\1\ DIRECTOR OF INFORMATION
TECHNOLOGY ACQUISITION MANAGEMENT ISSUES, U.S. GOVERNMENT
ACCOUNTABILITY OFFICE
Ms. Cha. I would like to express my thanks to you, Senator
Lankford, to Chairman Johnson, Ranking Member Carper, and
Members of the Committee for inviting me to testify today.
---------------------------------------------------------------------------
\1\ The joint prepared statement of Ms. Cha and Mr. Goldenkoff
appears in the Appendix on page 45.
---------------------------------------------------------------------------
IT implementation will be a key factor in whether the
Bureau will be able to adequately contain the costs for 2020.
The Bureau's past efforts have not always gone well. Our work
on the 2010 Census highlighted the mismanagement and major
cost, schedule, and performance issues associated with several
critical IT investments, one of which was intended as a cost-
savings measure but instead increased the costs of the Census
by up to $3 billion.
For 2020, the Bureau will rely on an enterprisewide IT
initiative called ``CEDCaP'' to deliver the systems and IT
infrastructure needed to carry out its cost-savings
initiatives.
For example, CEDCaP is planning to deliver the online
survey instrument and a cloud computing solution to support an
Internet response option. For field reengineering, the program
is planning to implement a new system to track and manage field
work. It will also test the use of mobile devices, either
government-issued or employee-owned, for field data collection.
Recent estimates put the program's cost at about $548
million through 2020. Given the Bureau's prior and existing
challenges, we highlighted CEDCaP as part of a new entry onto
this year's GAO high-risk list as one of a handful of major IT
investments in need of the most attention, and we plan to
initiate an in-depth review for this Committee later this year
on CEDCaP.
Based on our work to date, I would like to highlight two
key challenges this afternoon regarding the Bureau's IT plans
for 2020.
First, the time constraints. September's decision is
expected to drive the business requirements for CEDCaP's
systems and infrastructure. This milestone, which has already
been delayed by a year, cannot afford to slip further. However,
as we reported earlier this year, the Bureau had not yet
addressed how two critical inputs into this decision would be
addressed. These inputs relate to the Internet self-response
rate and the IT infrastructure security and scalability needs.
And if they are not adequately addressed by September, it could
lead to system rework downstream, eating into an already narrow
schedule margin.
By October 2018, the Bureau intends to begin end-to-end
testing to validate that CEDCaP's systems are ready to go live
on Census Day. This gives the Bureau roughly 3\1/2\ years to
develop and integrate planned systems, which may seem like a
lot of time but is not based on past performance, and a lack of
experience implementing technologies at the scale of the 2020.
The October 2018 milestone date is a prudent one for
planning purposes. If this date slips to the right and system
testing is compressed, then we very well could have another
HealthCare.gov on our hands.
In order to decrease the risk of system rework downstream
and protect the October testing date, the Bureau should fully
implement the recommendations we made to ensure that this
September's decision will be adequately informed.
The second key challenge is the Bureau's current IT
posture. The Bureau has made important progress to strengthen
and institutionalize IT governance and requirements management,
and as a result, the Bureau is better positioned to manage and
oversee its IT investments than in prior decennials. However,
more work is needed to address critical IT workforce gaps and
information security deficiencies.
In particular, the Bureau does not have the requisite
competencies in enough numbers to meet IT workforce needs.
These skills gaps include systems engineering, IT security,
cloud computing, and Internet data collection, to name a few.
To the Bureau's credit, it is aggressively working to close
these gaps. However, if they do remain open, the Bureau's
ability to deliver CEDCaP effectively will be hampered.
In addition, the Bureau does not yet have a comprehensive
information security program, something we first reported on in
January 2013. Among other things, the Bureau had not
effectively implemented appropriate access controls to protect
its systems from intrusion. For example, the Bureau did not
adequately encrypt Title 13 data in transmission and at rest,
nor did it use secure protocols to manage its IT
infrastructure, which placed sensitive data such as
administrative user accounts and passwords at risk of
compromise. As a result, we made 115 recommendations to address
these control deficiencies. As of today, the Bureau has fully
addressed 19 of them.
Given that the Bureau is considering using IT systems to
collect the public's personal information in ways that have not
been used in the prior decennial, implementing our security
recommendations must be a high priority.
In summary, while the Bureau is in a stronger position to
manage its IT investments than in the 2010 Census, it is also
starting development work much later in the decade. The margin
for error is slim. And the current workforce and information
security gaps add significant risk to the 2020 Census. Moving
forward, swift actions to fully implement our open
recommendations must be taken. Doing so will improve the
Bureau's ability to deliver on its IT plans and realize cost
savings.
That concludes my statement, and I look forward to
addressing your questions.
Senator Lankford. Thank you, all of you.
And as promised, Senator Carper is here and will give an
opening statement.
OPENING STATEMENT OF SENATOR CARPER
Senator Carper. Thanks so much, Mr. Chairman. What I would
really like to do is get right into the questions. It is nice
of you to yield to me.
I would say that I would like to ask that my statement be
made part of the record,\1\ unless one of my colleagues
objects.
---------------------------------------------------------------------------
\1\ The prepared statement of Senator Carper appears in the
Appendix on page 35.
---------------------------------------------------------------------------
Senator Lankford. Without objection.
Senator Carper. All right. Good.
Just very briefly, we have been doing this for 200-and-some
years--not us here. We have not been in the Senate that long.
But we have had Senators interested in trying to get a good
Census and trying to do it in a cost-effective way.
A lot of people think that we in Washington cannot organize
and run a good three-car funeral procession. And we have a
great opportunity here to demonstrate, by careful planning,
smart funding, good oversight, that we can do some amazing
things with the technology and the tools and the people that we
have. It is just really imperative that we deliver. And this
oversight hearing today is in part designed to make sure that
you can deliver and that we will be proud of the work that is
done in the next 5 years.
Thank you.
Senator Lankford. Thank you.
Let me address just a few questions, and then we will
continue moving through questions here on the dais as people
arrived.
Mr. Thompson, you had mentioned that there are some things
you need to talk to Congress about, about vacant units. Is
there something you are needing from Congress directly,
clarification or statutory changes?
Mr. Thompson. Thank you, Senator. There is, but let me
clarify my statement first. So in collecting the information
from our non-responding households to the Census, we are
testing ways in which we can use administrative records to
enumerate occupied housing units instead of direct in-person
enumeration. And we want to make sure that we are on the same
page as the Congress in terms of taking this step, if we should
propose to take this step.
Now, with regard to help from the Congress, there is a data
set--it is called the National Directory of New Hires (NDNH)--
and it is maintained at the Department of Health and Human
Services (HHS), and it would help us greatly in our program,
and we would need some legislative changes to have access to
that file.
Senator Lankford. OK. What other data do you need here?
Obviously there are multiple different databases between the
IRS, Social Security, Department of Labor (DOL); HHS has it,
you have it. There are lots of data sets around there. I guess
the question for me initially would be: Which particular data
sets do you want to be able to use? And, second, are any of
these agencies charging Commerce to be able to have access to
that data?
Mr. Thompson. Thank you. So you have named some of the data
sets that we already have access to. We are also in the process
of acquiring the Supplemental Nutrition Assistance Program
(SNAP) records State by State and the Women, Infants, and
Children (WIC) program records so we can put these together in
a very private way, protecting privacy, and using them and
research them as we proceed with the Census.
We are not being charged by the other agencies for these
data sets. In fact, the Secretary of Commerce has great
authority to ask for these records in conducting the job of
carrying out the Census.
Senator Lankford. OK. We had an extensive conversation here
at this same Committee room just a couple of weeks ago about
the Death Master File (DMF), and that we have about 6.5 million
people that are listed in our systems right now that are over
114 years old, I believe--112? Excuse me. A lot younger than
that, 112 years old, when we actually think there are less than
two dozen in the United States. So we have some problems in
those data sets that are out there. How are you going to try to
address that to make sure we do not get individuals into the
system that are not really alive anymore?
Mr. Thompson. Part of our research program, is to look at
which records can be effective and in which combinations of
records that link together. So we do need to do an extensive
research program to feel comfortable before we use
administrative records. And we are looking at those very things
right now.
Senator Lankford. OK. That is one of the aspects that we
will need to resolve, both how that is resolved within Census
and then obviously we would like to know that information so we
can get a chance to share that with other entities within the
Federal Government, because as I mentioned, with the Death
Master File, that has extensive millions of names on it that
are no longer with us anymore. We have to be able to resolve
that.
Mr. Goldenkoff, you have a very interesting statement,
which I have read and seen multiple times before. If we can put
all of these different aspects into place, all these different
innovations, we could save $5 billion. And so by saving $5
billion, we would spend the same as we did last time.
Now, if I sold you a pizza for $10 and said I am going to
sell you the next one for--I am going to save $5 for you, and
it is going to be the same cost, I am not sure you would think
that was a savings. So I am working through the math here
between how we save $5 billion, yet we do it for the same cost
as we did last time.
Mr. Goldenkoff. Well, part of it is a unit cost, roughly
$100 per housing unit. So you would expect to see some cost
increase because the population----
Senator Lankford. Right, because of the number of people.
Mr. Goldenkoff. Exactly. The workload is more. So that is
why we look at the unit cost, and it works out roughly the
same, about $100 per housing unit.
Senator Lankford. Is there a way to be able to save unit
cost, that the unit cost goes down? Or are we tapped out? I
think it is $97, the last stat that I saw, per person to be
able to do this. Is there a way to be able to save money per
unit cost rather than saying we had a growing population, we
assume it still costs $97 per person to do this.
Mr. Goldenkoff. Well, there is always a cost-quality
tradeoff. You can always conduct a count for less. The question
is: How accurate is it going to be? And that is always--that
has to be weighed. I mean, the Constitution requires that
everybody be counted, and so that is the gold standard. And so
because of the increasing complexities of counting everybody,
the cost goes up from Census to Census. The Census Bureau needs
to work harder each decade just to stay in place.
Senator Lankford. So going back to something Ms. Cha
mentioned before, in 2010 we had a $3 billion process of
innovation that came through. I think that was dealing with the
handhelds--is that correct?--is what it circled around.
Ms. Cha. That is correct.
Senator Lankford. So a $3 billion cost. A $13 billion
program really cost $10 billion because we had a $3 billion
program that ended up not being fully implemented and we lost
there. So what I am trying to figure out is we save $5 billion,
however we are still at basically the same cost as we did last
time per person, and we had a $3 billion failure last time.
Mr. Goldenkoff. That is correct. And I also need to say
that, GAO has not verified these cost estimates, so we are
relying for now on the Bureau's estimates. We will probably be
doing work in the future where we will look into the validity
of those estimates.
Senator Lankford. Is there a difference in the American
Community Survey and the cost per person in that piece versus
the every-10-year-Census piece, cost per person?
Mr. Goldenkoff. I do not have that information.
Senator Lankford. Mr. Thompson, do you happen to know that
number, cost per person, the American Community Survey that we
do every quarter basically?
Mr. Thompson. I am sorry. I do not have that figure.
Senator Lankford. Do you think it is higher or lower? Just
a ballpark on that.
Mr. Thompson. I think it is probably higher for the
American Community Survey, somewhat.
Senator Lankford. OK. I am going to keep going on
questions, and I will come around for a second round here in
just a moment. I will recognize Senator Carper, the Ranking
Member, for the next set of questions.
Senator Carper. Why don't I just go last? I would like to
hear the others.
Senator Lankford. OK.
Senator Carper. Then I will just come in at the end. Thank
you.
Senator Lankford. Senator Ayotte, you came in next after
that.
OPENING STATEMENT OF SENATOR AYOTTE
Senator Ayotte. Thank you. Thank you all for being here.
Here is what I wanted to understand. I know that you have
taken steps to use commercial mapping information in putting
together the proposal for the upcoming Census. And as I
understand it, you are using portions of that commercial data
to fill in gaps, but you still would be creating essentially
your own product of the map that you need for this. What I
wanted to understand is why we do not further use the
commercial products that may be available. Why is the Request
for Proposal (RFP) limited to filling in gaps rather than
relying more heavily on commercial, already existing products
where they already have? I know that their products may not be
able to fully do this, but could do a significant amount of it.
So what is the thinking there?
Mr. Thompson. Well, thank you, Senator. We have been
talking with a number of the companies in that field. We have
an RFP, as you noticed.
Senator Ayotte. Right.
Mr. Thompson. We want to evaluate the quality of the
materials that are available in the private sector and where
they will work and save us from walking the ground where we
intend to use them.
We also will be having another RFP coming out soon which is
going to be asking the private sector for services such as
automated change detection and the like.
So we are looking forward to seeing the responses to the
RFP, and we are looking forward to using commercial data where
it is available.
Senator Ayotte. So couldn't commercial data be used in a
much broader capacity than what your current RFP suggests?
Mr. Thompson. I did not think that we were trying to limit
the commercial data through the RFP process.
Senator Ayotte. So are you open to--for example, if
commercial data for mapping would get you--I am just using a
number--90 percent there and you had to fill in the rest to
complete your product, are you open to doing that?
Mr. Thompson. Yes.
Senator Ayotte. OK.
Mr. Thompson. Yes.
Senator Ayotte. It seems to me, I mean, so many of us use
this data in so many capacities, and so always re-creating the
wheel, this is a place where I know you are already taking
important steps to do this, but we could save a lot of money
for the public.
Mr. Thompson. Yes.
Senator Ayotte. I wanted to follow-up on some of the GAO
findings. What are the issues that you are going to be able
to--one of the big things, I think, that overlies all of this
is transparency in costs and measuring and achieving savings.
How do you think that you will meet the transparency goals so
that we can actually understand how much we are paying for
things in a much more open way so that when we have our
oversight function we do not end up in a position where we are
not having the right type of oversight?
Mr. Thompson. Well, thank you. We are committed for this
Census and the 2020 cycle to being very open and transparent
with our methods. I think we have a good relationship with the
GAO. We certainly hope that they will--and we know they will--
look carefully at what we are doing.
I should also note that once a quarter we have a project
management review in a public way for the 2020 program, and we
webcast it. We make it available and we want everyone to see
how we are planning and what we are doing, and we are trying to
be very transparent with that.
Senator Ayotte. So, Ms. Cha, thank you for your testimony,
and I think it is an eye opener for all of us when we hear
potential analogies to HealthCare.gov, because obviously we all
got an earful from our constituents and others over that
rollout.
What are the most significant steps that we could take in
this Committee to make sure that, in fact, this IT procurement
goes properly? One of the issues that we have I think overall
in the government--it is not just unique to this. We are
particularly bad at this. Throughout the government we have had
numerous occasions where we have invested a lot in a particular
system, and we have not ended up with the system that we hoped
for or the costs went way over. So I just wanted to hear from
your perspective. If you were sitting in our shoes, what would
you think that we should do most to make sure that the warnings
that you have given us do not occur?
Ms. Cha. Well, Senator, your continued and sustained
oversight of the 2020 Census and particularly the CEDCaP
program will be vital. These large and complex IT modernization
programs are challenging, and it is one of the reasons why we
cited CEDCaP as part of a new GAO high-risk entry this year in
terms of improving IT management governmentwide. These programs
rarely meet cost schedule and performance goals, and we have an
opportunity here. The train has not left the station. And so,
when we look at CEDCaP and we look at the complexity and the
risks associated with this program and you layer on top of that
the key challenges that I identified in terms of the time
constraints as well as the risks associated with the IT
workforce gap and the information security control
deficiencies, there is quite a bit of risk here. I think there
is an imbalance.
And so to the extent that that the Bureau can focus on
implementing a simpler solution, that would, I think, set the
Bureau up for success relative to IT.
Senator Ayotte. So as I see it, I hear you saying that the
solution they are proposing may be too complex to achieve
during this period with the workforce that they have.
Ms. Cha. That is correct.
Senator Ayotte. And is there anything more we should be
doing also on the workforce gap issue? Which, frankly, is
something that we face across government agencies, because this
is a highly competitive field.
Ms. Cha. Well, again, it is that continued monitoring on
your end. I do want to say that the Bureau is aggressively
looking to close these gaps that they do have. However, if
these gaps are not effectively closed, then I think the Bureau
needs to, again, continue to identify ways to decrease the
complexity of the CEDCaP initiative.
Senator Ayotte. So my timing is up, but I would like to
hear what your response would be to that about making this a
simpler process so it could be easier to achieve.
Mr. Thompson. Certainly, Senator. So I think I would coin a
phrase from our Chief Information Officer (CIO), and it is
something that worked very successfully in the 2000 Census when
we did deliver our systems on time and on schedule, and that
is, we are innovating but we are not inventing. So our plan is
to use existing technology and existing expertise and innovate
by using that intelligently.
So, for example, we are going to be using a smartphone, but
the smartphone already exists. We are not going to invent one
like we tried to invent a handheld the last time.
We are using optimized systems for our routing, but we are
not building--those systems already exist, and a number of
people have very good ones, including the United Parcel Service
(UPS). So we are trying to take advantage of what is existing
and not trying to invent new things.
Senator Ayotte. Thank you.
Senator Lankford. Senator Peters.
OPENING STATEMENT OF SENATOR PETERS
Senator Peters. Thank you for being here and for your
testimony. As Senator Carper mentioned, it is important to get
this right, and certainly it is a monumental undertaking to be
able to count every individual in this country, and this is not
an easy task. I appreciate that. But it is a very important
task that you undertake, particularly from a public policy
standpoint, so that we can make sure folks are protected, have
resources, allocations that are done properly, so my hat goes
off to you.
I represent an area in Michigan, however, that has a very
large population of folks from the Middle East and from North
Africa. Mr. Thompson, I sent a letter to you earlier this year
with a couple of my colleagues talking about that issue and the
fact that that classification is not part of the Census. And as
you know, it can be problematic because if you are not on the
Census, you do not have access to some voting rights
protections, including access to ballots in your home language
to be able to vote and exercise that right. Also, it is
difficult for researchers if we do not have that type of Census
information to understand potential health disparities that may
exist in the Middle Eastern population in Michigan and other
places around the country, as well as employment discrimination
and a whole host of things that we need to look into, and this
would provide accurate data.
My understanding is that you are going to test that
category for Middle Eastern descent as well as North African. I
appreciate your prompt response to our letter a ways back, but
I want to take this opportunity to get an update from you as to
where you are. I know when you responded to my letter, you
mentioned you were going to be reaching out to stakeholders in
the spring and consulting with community leaders within that
community to get some assessment as to how you would proceed.
Could you give me an update, please, on where we stand on that
and how you look to proceed?
Mr. Thompson. Certainly. So we have been having ongoing
discussions with various members of the Middle Eastern and
North African community, and we are testing including that
categorization as part of our race and ethnic questions, and
that test will take place this fall.
We are looking forward to having an expert meeting of
various Middle Eastern and North African scholars this spring
so that we can really come up with a definition, if you will,
of Middle Eastern and North African that everyone agrees to so
we can use that as part of how we explain it when we conduct
the Census. So we are on track.
Senator Peters. Well, that is good to hear. I appreciate
that.
Have you heard from any other Americans or other
organizations about a lack of representation in the Census? Is
this unique to Middle Eastern community, or are there others
that you are working on?
Mr. Thompson. In terms of the----
Senator Peters. Classification.
Mr. Thompson. Classification of Middle Eastern and North
African. We have heard from other constituents on other aspects
of the question, but for inclusion, it has been basically the
Middle Eastern and North African.
Senator Peters. So that is where your focus is, it is going
to be on this going forward?
Mr. Thompson. Yes.
Senator Peters. Great. I have many constituents, as you can
imagine, given the large size of our community in Michigan,
that are very passionate about this, and I understand that you
have received thousands of inquiries from the community. Have
you been responding directly? Or how can I respond to my
constituents as to how you are listening to them and the
response that they will receive?
Mr. Thompson. I think what you are referring to is a
Federal Register notice we put out that announced our plans for
this fall test, and we received well over a thousand
recommendations that we include in that test the Middle Eastern
and North African category. And we are responding by a comment
to the Federal Register notice to those recommendations that we
got.
Senator Peters. OK.
Mr. Thompson. Now, we will group them together because many
recommendations said the same thing.
Senator Peters. Right. So you will be summarizing that in
the Register.
Mr. Thompson. Yes.
Senator Peters. And the folks we could refer them to that
as well as get back----
Mr. Thompson. Yes.
Senator Peters. Well, I appreciate that. The other issue
that I am concerned about deals with the homeless population. I
know you are reaching out to folks who are not going to be at
an address to count to make sure that they are represented. In
fact, in Michigan, Washtenaw County is one of 71 communities
that participates in Zero: 2016, which is a campaign to end
veteran and chronic homelessness. And as a result of a recent
survey that they did, they did show a 24-percent decline in the
number of people living on the streets and in shelters, which
is a good thing. But I am particularly concerned, as they are,
with homeless veterans who may be out there, who have the
ability and certainly the right to access benefits, but may not
be aware of it.
It seems to me having Census workers out finding people who
are living on the street--and you have the staffing to do
that--presents an opportunity, first, to identify those
individuals, but also, two, present them with information that
may be helpful to get them to a place where they can get the
services that they have earned and they deserve.
Could you talk a little bit about how you reach out to
folks who are homeless living on the streets and if you would
be in a position to help in our efforts to bring these folks
out of the shadows and into the help that they deserve and
need?
Mr. Thompson. Senator, getting an accurate count of all
Americans is our primary goal for 2020. We are right now at
this point in the process of starting to look at how we will
blend in operations on top of our automation and reengineering
to count these populations. So, for example,
I have a National Advisory Committee on Racial, Ethnic, and
Other Populations, and we are starting a working group on that
committee right now to start addressing these issues.
We want to address these issues over the years ahead, but
right now I do not have a proposal for how we are going to
count the homeless and the veterans in 2020. But we do intend
to have a plan for that, and we will share it widely.
Senator Peters. When do you expect to have that plan?
Mr. Thompson. I would have to get back to you. I would say
by the end of next year we should have some significant
progress on that.
Senator Peters. Well, I would appreciate that, and if you
could inform our office about that, I would appreciate that.
And then the second step is that when we identify these
individuals, it also is an opportunity to make sure that these
individuals, particularly homeless veterans, that there are
benefits that are available to them and places of refuge for
them to go. Although it is a different mission than you have
for your Census workers now because of certainly the concern
about those who have served our country who may be homeless,
this may be an opportunity for us to kind of take a whole-
government approach to helping those individuals. I would hope
you would be open to that as well.
Mr. Thompson. Of course.
Senator Peters. Thank you.
Senator Lankford. Mr. Thompson, before I recognize Senator
Carper, can I get a clarification? By the end of 2015 or by the
end of 2016 you will have the plan for the homeless?
Mr. Thompson. I was referring to the end of 2016.
Senator Lankford. OK. So by the end of 2016, you will be
able to have articulated a plan for that, for identifying the
homeless and then doing a count there?
Mr. Thompson. Yes.
Senator Lankford. Great. Thank you. Senator Carper.
Senator Carper. Thank you, Mr. Chairman.
Mr. Thompson, it is very nice to see you, and Mr.
Goldenkoff, Ms. Cha. I love saying your name.
This is not your first opportunity to be involved at a
senior level for the Census. Is that correct, Mr. Thompson?
Mr. Thompson. That is correct, Senator. I was the career
person in charge of the 2000 Census.
Senator Carper. OK. Just thinking back on your tenure with
the Census Bureau, share with us some of the insights or things
you have learned in this tour as our Director that are new and
that you did not know before and how might they direct us as we
try to get a better result for less money?
Mr. Thompson. Well, Senator, you are hitting on the reason
that I came back into public service. When I left the
government in 2002 and went to the private sector, I had no
intention of coming back into government. But I could not
foresee just the way that technology had developed and offered
a lot of opportunities to do the Census fundamentally
differently than we had done it before. And that got me excited
to come back into government and help the Census Bureau move
forward.
Senator Carper. When you watched back in 2010, or actually
the run-up to 2010, when they were trying to develop the
handheld, and rather than get something off the shelf, actually
develop and build and invent it, as you said, invent it
ourselves, what did you think as you watched that unfold?
Mr. Thompson. I was very disappointed that that happened,
that that event happened. I was fairly close to it. I was on a
panel that Secretary Carlos Gutierrez called together to give
him a recommendation on how to proceed in the aftermath of the
handheld contract issues. I was just disappointed and sorry for
the Census Bureau. I wish I could have been there, but I was
not.
Senator Carper. My recollection was this is when Dr. Coburn
and I were taking turns leading the Subcommittee on Federal
Financial Management that had jurisdiction over the Census. And
it seems to me we went through a period of time where we had a
fair amount of churning in terms of leadership at the Census.
Is that correct? I think we had someone who was Acting
Director, and then within the span of a couple years, we had
maybe a couple of other people who were Acting Directors. We do
not have that challenge anymore, do we?
Mr. Thompson. Not right now.
Senator Carper. That is right.
Mr. Thompson. I am delighted to be the Director.
Senator Carper. Because we have a law that actually says
that the folks who are appointed in the job serve for 5 years,
and the idea is that hopefully we will end up in a situation
where we are not going to be in the throes or we are in the
face of the Census and all of a sudden have turnover in
leadership. My hope is that is going to serve us well.
I am a guy who likes to think about incentives, how do we
incentivize behavior to get better results, and I am real
interested in root causes. We have a lot of people that are not
interested in cooperating and being counted and providing
information to the Census because of their right to privacy or
they do not see any value for taking the time to either respond
online or in person if somebody knocks on their door. They do
not have an appreciation for the value of the information that
we gain from the Census. They do not care how their
Congressional districts are apportioned and all that.
How do we incentivize people to be better citizens in this
process? What are we doing, we collectively, to incentivize
people to be better citizens in this regard? It is all great
that we have the technology, we provide the money, the
oversight, the funding. But how do we incentivize people?
Collectively, how do we do that?
Mr. Thompson. That is a very good question, Senator. We
have started research on that issue already for the 2020
Census. We have a test in the Savannah, Georgia, area, the
Savannah, Georgia, media market, and we have already started to
test different ways to reach out to the American public to
encourage them to respond. And this is just the first in a
series of tests, but we are starting this process much earlier
than we did in the 2010 Census process.
Senator Carper. OK.
Mr. Thompson. But it is something that we have to conduct
some research on because the population has changed. It is not
the same population that we had in 2010.
Senator Carper. All right. I agree with you. I thought that
was a pretty good question, and I am going to ask you to think
about your answer and respond to me on the record, because I
think this is a big issue, and most people have not a clue of
the value to our country, to their communities, of a Census
done well.
Let me also ask, one of my many favorite sayings is, ``Pay
me now or pay me later.'' And I think that also applies with
the President's budget request for 2016, and he is asking in
his budget for the Census a bump-up in funding. Explain to us
why that relates to pay me now or pay me later.
Mr. Thompson. Certainly. So the biggest part of the bump-up
is to start developing the systems and processes that we will
use to take this reengineered Census, and it is a challenge
because we have never done a Census this way before. We have
never used handheld devices. We have never used the Internet as
a self-response option for the Census. We have never used
optimized routing. So there are a lot of gains through that.
But we really have to do the work in 2016 and 2017 to have in
place the system so that we can run a complete end-to-end test
in 2018, and that would give us the assurance that we could
avoid a bad situation, similar to what happened before.
Senator Carper. OK. I have some more questions, and I will
look forward to another round. It will give me a chance to pick
on our other two witnesses. Do not go away.
Senator Lankford. Thank you. We will open it up for
questions here, and we will get a chance to go through as much
as we can.
Mr. Thompson, you had made a comment about identifiers and
the Internet, as far as being able to follow-up. Can you
complete that thought as well? If individuals go on to the
Internet and they actually do their information, will they have
an identifier there that is consistent so we will know who that
was and have the consistency of that, as we would with a paper
form?
Mr. Thompson. Thank you, Senator. So let me be clear. What
we are testing is in every Census to date, most of the
enumeration required that an individual have either a Census
questionnaire or an ID number that linked them to an address.
What we are testing now is letting people respond with their
address alone, and that has two issues that we have to feel
comfortable with before we do it. One issue is we have to be
able to validate their response as accurate. And the other
issue is we know that in the past even we have had more than
one response for the same address. And so we are going to have
to be able to unduplicate the responses that we get. These are
some challenging problems. We are working with experts in the
field, including the National Institute of Standards and
Technology, the MITRE Corporation, and some others.
So we are taking this very seriously, and we need to be
convinced that it will work. But if we can make it work, it
does give us the opportunity to allow people to be counted when
we reach them with a message if they do not have their
questionnaire. So we think that the opportunity there is worth
trying to do that. But as I said, we are testing it. We are
going to have the results of our testing, and we are only going
to implement it if we feel comfortable.
Senator Lankford. So the assumption is that we would have a
master list of every address in the United States, and if they
pull up that particular address, they could fill out the
information for that address, not that they are creating an
address in the system. So there is some sort of verification
even at the beginning if they are typing in an identifying
address that we do not recognize as a real address.
Mr. Thompson. Yes.
Senator Lankford. OK. Ms. Cha, questions and thoughts about
that? What are the risks involved?
Ms. Cha. Well, in addition to the address verification,
there is also the identity verification as well. So I just
wanted to kind of round out that piece, these two key elements.
On the former, it seems to be much more simple from a
technology perspective to match the address that a respondent
enters into the form and match that against the Master Address
File (MAF). However, the concern that we have here is that it
is unclear as to whether the Bureau knows exactly what it needs
to know in time for its September 2015 decision. Again, that
milestone decision is going to--it is intended to deliver the
complete set of business requirements for the systems
development piece.
So if there is information regarding non-ID processing that
is not contained in that decision, then ultimately it could
lead to system rework downstream. So that is the primary
concern that we have with regard to non-ID processing, just
that the Bureau is--we are in year 5 at this time, and if
continued testing in this area proceeds beyond September 2015,
it will impact the systems that they are expecting to deliver
and to deploy.
Senator Lankford. Right. And you made extensive comments
about cybersecurity issues and vulnerabilities. Your statement,
if I remember correctly, access controls, you have 115
recommendations of which 19 have actually been responded to at
this point.
Ms. Cha. That is correct.
Senator Lankford. So what does that mean as far as your
perception? At this point you would expect 115, 19 have been
responded to. When would you anticipate the other 90-some-odd
are going to be responded to?
Ms. Cha. Well, we would like to see this be one of the top
priorities for the Census Bureau. My understanding from my team
is that the Bureau provided additional evidence late Friday
afternoon, so my team will be looking at the progress that they
have made in the remaining recommendations. And we are
committed to working with the Bureau to ensure that they
effectively close those out, because ensuring that the
information that they are collecting is secured is of vital
importance.
Senator Lankford. Sure. Mr. Thompson, do you want to
comment on any of that at this point?
Mr. Thompson. Yes, Senator. My team is of the perspective
that we have responded to all of the concerns. However, we have
obviously not communicated that to our colleagues at the GAO,
and I certainly intend to make sure we communicate that and we
reach agreement as to which ones have been responded to, which
ones have not, and what we have to do. And then I would be
pleased to send a report back to you.
Senator Lankford. Good. There has been great conversation
in the last several years about the questions that are included
or not included. There was a little bit of conversation in 2010
over the issue of citizenship, as I am sure you will remember
distinctly. Where do we stand on that conversation for 2020 as
far as whether citizenship will be asked or are we going to get
every single house, and then other questions that may change or
that are being currently considered and when that list will get
to us and we are evaluating these questions to be different
than the previous 10 that were there, and when will we get a
chance to take a look at those as well?
Mr. Thompson. Senator, right now the citizenship question
is on the American Community Survey. We have not received a
request, because that is what determines the content of the
Census, from those entities that have asked us to put that on
there. So we have not received a request to put citizenship on
the Census itself.
Let me be clear about a couple things. In 2017, we will be
submitting the topics that we plan to ask on both the Census
and the American Community Survey to the Congress, and in 2018,
we will be submitting the actual questions.
Senator Lankford. OK. And let me ask about the American
Community Survey. There has been a lot of consternation in
Oklahoma, and I would assume in other States as well, about the
nature of how it is collected if you do not respond to it. My
office consistently gets calls from individuals saying someone
is at their door, someone is parked outside. We have a single
mom that is very uncomfortable that there is some guy outside
in a car; they are knocking on the door consistently. All of
those things to get the American Community Survey. How are we
going to put this into a process that they understand this is a
Federal Government that is working for them rather than a
Federal Government that is stalking them? If they do not want
to complete this, shifting that to a voluntary status, and so
we can have some way that they do not feel like it is
oppressive coming after them to be able to fill out the
American Community Survey?
Mr. Thompson. Thank you, Senator. I have heard the concerns
about the American Community Survey. While it provides a wealth
of very important information that is used by the Federal
Government to allocate funds by State and local governments and
by businesses to create jobs, we understand that there are
concerns with the length of the American Community Survey and
with the perception that the language on the envelope which
says your response is required by law is offsetting. So we have
embarked on a program to look at how we can address these
concerns, and we are doing a number of things. We are looking
to see if we can make the questionnaire shorter by asking some
questions every other year if the data is not needed on an
annual basis.
Senator Lankford. There is some publicly available data as
well that we seem to replicate.
Mr. Thompson. Right. And we are also looking to see if we
can use administrative data records to answer some of the
questions on the form. We are looking at how we can increase
our messaging, change our messaging to be more of a
communicator, why it is so important that you fill it out, what
it means to your community. And we are also beginning a program
to test the language that we put on the envelope. So this May,
we are running the test. That is next month. We are doing a
test to measure the effect of not putting that language on the
envelope so we can see what the effect is. So we are taking
these concerns very seriously. The Secretary of Commerce is
very attuned to this. I am providing a report to her at the end
of this fiscal year, and she expects to see some significant
progress in this area.
Senator Lankford. Well, surely you can understand someone
that shows up on your doorstep repetitively with a badge
saying, ``I would like to ask you: What time do you leave for
work? How many people are in your house? How many toilets do
you have in your house?'' those questions immediately raise red
flags for a lot of Americans. They do not want to answer it.
But they also cannot make this person go away, which is why our
office gets calls to say, ``Why is this person here? Why won't
they go away? I have told them no, but yet they just keep
coming at me.''
So it is not just the nature of the questions, the length
of the questions, the type of questions. It is the treatment of
the individual that is at their door and the repetitive push
for them. So there is a whole series of issues there. I
understand the information is valuable, but there are a lot of
issues that are still there in the American Community Survey
that we have to resolve, and I am glad to hear that--you said
this testing is in May. When will we get results from this?
Mr. Thompson. I do not know exactly, but we do not intend
to take that long to get it out.
Senator Lankford. OK. So give me an example: By December?
Mr. Thompson. Oh, definitely by December.
Senator Lankford. OK. So this year, this summer? Into the
summer?
Mr. Thompson. I would hope so. But I do not want to commit
to it, but we will definitely have it by December, probably
much sooner than that.
Senator Lankford. OK. Thank you. Senator Carper.
Senator Carper. Before I move away from the American
Community Survey, let us say if my household were one of those
chosen in Delaware for the completion of the survey, is the
first notification that we have been selected somebody knocking
on my door? What would precede that?
Mr. Thompson. The first thing you would see would be a
postcard notifying you that you have been selected for the
survey in the mail.
Senator Carper. And then what?
Mr. Thompson. And then you would get a letter asking you to
respond by the Internet, and then you would get a questionnaire
asking you to fill out the questionnaire and mail it back.
At that point the next step would be if we had a telephone
number for you, we would call the telephone number. The next
step would be then we would send a person to visit you in
person.
Senator Carper. OK. Is there any time during that process
we somehow convey to people why this is important for them to
do it, to respond?
Mr. Thompson. Senator, we do do that, but I think that
there is room for us to improve on that. For example, we have
already had one training of our entire field staff on how to
approach households, and we are going to be doing additional
trainings in this area. We need to do a better job of
communicating to the American public why this survey is so
valuable to them and why----
Senator Carper. Just to keep in mind maybe what I hope we
already do this, but whether it is in the postcard or whether
it is in something that folks get over the Internet or
something that people get in writing saying, ``You are going to
get a survey,'' just to be able to say in a very short few
words why this is important, without a whole lot of verbiage.
I want to sort of refocus for a little bit, if we could,
and talk about cybersecurity. Ms. Cha, we are going to get you
in the act here. Let me just ask, what do you believe needs to
be done with respect to cybersecurity, on the cybersecurity
front, as we prepare for the 2020 Census?
Ms. Cha. First and foremost, it has to be addressing the
115 recommendations that we have made relative to those control
deficiencies. Again, as I mentioned to Senator Lankford, access
controls in particular are of vital importance in
strengthening, because those regulate who and what is accessing
the Bureau's systems. And so, as the Bureau is exploring these
methods to collect data out in the field, including, using
personally owned devices or government-issued devices, it is
going to be critical for the Bureau to secure those back-end
systems to make sure that the information is properly secure.
In addition, if they are expanding the use of
administrative records, they are also going to have to shore up
the controls there to ensure that that information is
adequately protected.
So addressing our 115 recommendations made is the top
priority.
Senator Carper. 115? That sounds like a lot.
Ms. Cha. Well, actually, it is a little lower since the
Bureau at this time has addressed 19 of them, so roughly 20
percent.
Senator Carper. OK. You have already talked a little bit
about this today, I think, Mr. Thompson. Would you just kind of
think out loud again for us on the cyber front? As we sit here,
folks from all over the world are hacking into government
systems trying to steal our intellectual property, identifiable
information, and it is a huge challenge for us. What can you
tell us that would make us think that we are up to this
challenge?
Mr. Thompson. Yes. So we have a multi-layered system to
protect information. We limit the access to our systems. Within
our program we have ways to control who can access the data. We
encrypt all of our individual data when it is at rest, so, for
example, our philosophy is that if we have data on a machine,
we assume the machine can be compromised and, therefore, we
have to protect and encrypt the data.
We have some very sophisticated ways that we look at the
Internet traffic that moves in and out of the Census Bureau. We
work with the Department of Homeland Security on that
detection. We also scan all of our outgoing traffic to make
sure there is no personal information in it.
But having said that, and we do want to get straight with
the GAO on the 119 recommendations because we take that very
seriously, and we want to be in compliance with those because
that is the kind of practices that we need, and we will get to
the bottom of that.
Senator Carper. Let me just ask a question. Of the 115
minus 19--what is that, 96--outstanding, are some more
important than others? I presume they are.
Ms. Cha. Yes, and the access controls-related
recommendations are of the utmost priority.
Senator Carper. What Ms. Cha just said, is that something
that rings a bell, access controls?
Mr. Thompson. I believe we have addressed that. But, again,
I need to get with our people and her people and make sure that
what we think is addressed is what the GAO thinks is addressed,
because I would rely on them and their final say on that.
Senator Carper. OK. Let me ask you about hard-to-count
populations. We talked about this already, but it is certainly
one of the most challenging jobs in the Census and also one of
the most expensive, a real cost driver. Maybe what we learned
from the last couple of times out, 2000, 2010, about counting
hard-to-count populations, one of the ideas I floated, I guess
in the run-up to 2010, was maybe we could have--in terms of
incentivizing people, to find ways to really say, ``I would
love to''--``I cannot wait to respond to the Census because
this is what I will get out of it.'' We threw out ideas for
lotteries where people could have their name--if they were
early responders, they could be in a lottery, just all kinds of
ideas. But hard-to-count populations, people who do not want to
be counted, people who do not want you to know they are there,
folks who maybe because they are undocumented, or maybe because
they are parole violators, how do we get some of those people
to come out of the shadows?
Mr. Thompson. So one of the most important ways that we
have found to help count those populations is through a program
that we call ``local partnership,'' and that is where during
the decennial census we have a number of individuals that we
identified that have local ties to communities, they are
trusted voices in the community, and by working with these
people, giving them materials about why the Census is important
to their community in particular, not just in general but to
particular communities, we have seen gains in the 2000 Census
and the 2010 Census in reducing the differential undercount
between minority and non-minority populations. So we are seeing
those gains. It is not perfect, but it does show that the more
we can reach out on the local level with trusted voices to
convince people to be counted, that works and pays big
dividends. And that is something else we are starting to work
on at this point in the decade.
Senator Carper. Good. All right. Thanks. Thanks so much.
Mr. Chairman, if we have another round, I have a few more
questions.
Senator Lankford. Go ahead.
[Pause.]
Senator Carper. Mr. Goldenkoff and Ms. Cha, what is next
for GAO's oversight for the 2020 Census? Give us a look ahead.
Mr. Goldenkoff. Our focus is on the operational aspects and
the cost-savings estimates, so we are going to be looking at
these other four cost-savings initiatives that Director
Thompson mentioned. We have ongoing work that is looking at the
Census Bureau's use of administrative records. That report will
be out in a couple of months. After that, we will be looking at
the Census Bureau's use of targeted address canvassing, and we
also expect to look at the Census Bureau's cost estimates, and
we will try and validate them. Right now they are estimates,
but we do not know what is behind them.
Senator Carper. Ms. Cha.
Ms. Cha. In addition to that, Senator Carper, we will be
starting a review for this Committee with regards to the CEDCaP
initiative, and we will be starting that later this year,
roughly in the October timeframe. So that is going to be very
critical because we will know early on whether or not the
quality of the September 2015 decision is good or not in terms
of the completeness of the business requirements, again,
because having a full, comprehensive set of business
requirements is critical to informing the system development
work downstream.
Senator Carper. Why don't each of you give us--we have
already heard you in response to other questions talk a little
bit about this. But if you were sitting up here with us and you
wanted to make darn sure that we got a good Census at a
reasonable price, what are the one or two things that you would
do if you were in our shoes to make sure that that happened so
that it would not be on us and we would have better met our
responsibilities? Ms. Cha, why don't you go first? You look
like a potential Senator.
Ms. Cha. Well, I think going back to my response to Senator
Ayotte, with regards to the complexity of the CEDCaP initiative
itself, I think based on the risk that the Bureau currently has
relative to IT acquisition, it would behoove them to look at
whether there are simpler ways to go about acquiring the CEDCaP
initiative. It is quite ambitious, and, again----
Senator Carper. There are people who might be watching this
on television or listening, and they would say, ``What is
CEDCaP? '' Could you just explain it in terms that people might
understand?
Ms. Cha. Yes. It is the enterprisewide IT initiative that
will deliver the systems and the infrastructure to support the
2020 Census. So it is a collection of roughly 14 projects and
multiple systems beneath those projects. So it is a large and
complex IT modernization program, and it is critical that the
risks that we have identified with regards to the time
constraints as well as their IT posture in general, that those
risks inform the decisions about IT that they will be making in
September in terms of what is actually realistic and viable for
2020.
Senator Carper. All right. Same question. Mr. Goldenkoff,
please.
Mr. Goldenkoff. I would hold the Bureau accountable for
meeting costs and scheduled milestones. If you look at where
the 2000 Census went off the track and the 2010 Census went off
the track, past experience has shown that the Census Bureau
either runs out of time or runs out of money. So the best way
to make sure that the Census Bureau is on track is regularly
checking in with them and monitoring whether they are meeting
their cost milestones and their schedule milestones. And that
is the early indication that things are going off the rails.
Senator Carper. All right. Director Thompson, if you were
Senator Thompson and this was Director Lankford, what would you
make sure we get done?
Mr. Thompson. I would rather comment in a different way.
Senator Carper. OK.
Mr. Thompson. We are putting some things in place to try to
minimize the risk, because you can never be sure. So, for
example, the document we are producing this fall, the 2020
Census Operational Plan, and it will be 3 years earlier than we
issued an operational plan for the 2010 Census. So we are
getting an early start on it.
We have in place an active risk management framework (RMF)
and risk management process where we constantly identify major
risks, and we constantly look at how we can find contingencies
for those risks. We would be happy to share that with the
Committee so you can feel comfortable that we are looking at
the right risks.
In terms of the enterprise IT systems, we do not argue with
the GAO that it is a high-risk system because we are developing
new processes. But I can assure you that this is getting full
attention from the Department of Commerce. The Department of
Commerce CIO is fully engaged with this program and with our
CIO on the program. In fact, he is running an independent
verification and validation (IV&V) process on the program. So
he is taking it very seriously, too. The whole Department of
Commerce has this as one of their priorities, and success is
one of the priorities of the Department of Commerce.
Senator Carper. Good. Mr. Chairman, I have one more. Do we
have----
Senator Lankford. Sure.
Senator Carper. In my old job as Governor, when we were
wrestling with a problem in Delaware, trying to figure out how
to deal with a challenge, sometimes in a Cabinet meeting I
would say to my Cabinet, ``Some other Governor in some other
State has dealt with this issue. They figured out how to deal
with it in an appropriate way successfully. What we have to do
is find that State, that Governor, whoever worked on this issue
in another State, and find out: Did it really work? Is their
solution transferable to us in Delaware?''
We are not the only nation grappling with doing a Census,
and we are not the only nation grappling with doing it in a
more cost-effective way. Who are some other countries that we
are learning from? What are we learning from them?
Mr. Thompson. We regularly work with other countries like
the United Kingdom, Canada, Australia, and share our progress;
we share our problems. We have an ongoing dialogue about what
works, what does not work.
So, for example, Canada has been using the Internet as a
response option before we have, so we have been working with
the Statistics Canada people to understand what they have
learned that is both good and not good in terms of how to
implement an Internet self-response option. But we do have
active discussions going on with our fellow countries.
Senator Carper. Could you give us just maybe one example, a
clear, easy-to-understand example of something that we have
learned in this back-and-forth with other nations, maybe the
ones you mentioned, maybe others, say this is what we learned
from Canada, this is what we learned from Britain that we are
going to be doing differently this cycle?
Mr. Thompson. Well, I think there were two things with
Canada. One thing was that the Internet can be a very
successful way to get responses in on a very timely basis, and
Canada has had great success in using the Internet as a primary
response option.
The other thing we have learned from Canada is that if you
are not cautious about how you word the equivalent of the
American Community Survey and the messaging on that, you can
lose a lot of data quality. At the last minute Canada had to
make their survey voluntary in 2011, and they did not have much
time to work on it, and they lost a lot of data quality and
were not able to publish a lot of the data from their
equivalent to the American Community Survey. So those are two
examples of some things that we have learned.
Senator Carper. Good. Thanks.
Thanks, Mr. Chairman.
Senator Lankford. One other thing I would just ask you to
submit for the record is the controls and how you are going to
handle staff for 2020, background checks, citizenship status,
and such for actual Census takers and Census staff, so when
that is settled and resolved. Do you have a good idea of when
you are going to set those parameters?
Mr. Thompson. Right now, our intention is to use similar
processes in terms of fingerprinting and validation as we did
before. However, there is still an active class action suit on
that issue, and so we are awaiting the outcome of that, and I
do not have an estimate of when that suit may be resolved.
Senator Lankford. OK. So the assumption is to use basically
the same standards as the last time, pending this class action
suit.
Mr. Thompson. Yes.
Senator Lankford. OK. Then we will follow-up from there.
Going back to the American Community Survey again, is there
anything on the American Community Survey that is not publicly
available from other places or from other survey teams? I
understand our every-10-year Census, that one is
constitutionally mandated. The American Community Survey is
not. And so what I am looking for are some of the efficiencies
that are there of cost. What data can we buy at less
expensive--you are saying if it is $97 a person for a main
Census but you think it is more for the American Community
Survey, is there a company we can buy this same information
from for $50 apiece and save us half the money? That is what I
want to try to figure out and to see do we need to gather this
data this way from Census, or is there a more efficient way to
do that? And I would be interested to know if GAO has examined
that in the past.
So, Mr. Thompson, if you want to just be able to address
just some of the other resources that are out there that are
publicly available that may already be there. I am sure you are
talking about that with your questions, and then other places
and efficiencies.
Mr. Thompson. So, Senator, we are looking right now as we
speak at what resources are available to supplement or remove
some of the questions on the American Community Survey and
getting them from other places. I must say, though, having run
a business myself, my efforts in running the business were,
trying to keep the business in business and to grow the
business. And so I was not motivated to form a comprehensive
data series for the entire United States. I was looking at
where my data was most important to my business.
Senator Lankford. Right.
Mr. Thompson. And so that is why it would be very difficult
to get from the business community a comprehensive data set for
all parts of the United States.
That being said, we are certainly looking to use existing
information in lieu of asking people that information.
Senator Lankford. Right. So my question is: There are
businesses that do this kind of investigation all the time,
this kind of research. You are right, they are not trying to
comprehensively touch on every single area, but if we are able
to say to them, ``We want to buy your data. It has to include
this,'' they already have half the data, they could gather the
other half and probably half the cost for us as well. I am
trying to figure out how do we get below $100 apiece to be able
to ask a set of questions when everyone who does surveys--and
maybe I am missing it, but I do not know of another survey that
costs as much as that one does to be able to do it. I
understand the unique parameters that we have with every
household, every street. That is different. It is like U.S.
Mail has a different set of cost parameters than FedEx does
because it is every house, every day, every street. I get that.
But I am trying to find are there other ways. Has GAO examined
something like that in the past?
Mr. Goldenkoff. Not a very detailed analysis. We have not
gone question by question. What we have looked at is the
availability of some of the data in other Federal surveys and
other Federal data, and it really varies. In some cases there
really is no substitute for the ACS data because it really gets
data down to a very low level of geography, and sometimes that
is needed for some of the requirements. Some of it goes back to
statute. Some statutes require that the data come from the
Census. In other cases, it is just the best data available for
that need, and so it is more for administrative purposes.
So this is something that the Census Bureau has to look at,
go item by item and see what are the alternatives and what is
the cost-benefit, again, balancing that cost-accuracy tradeoff.
Senator Lankford. Right. But has GAO taken a look before
about private entities doing this if we paid them--again, if we
are at $100 a survey, are there entities that would gather the
same amount of data----
Mr. Goldenkoff. No, we have not----
Senator Lankford [continuing]. For less cost? Again, we
have a difference between the American Community Survey, the
Census between the Census, and the actual Census. We have a
constitutional mandate for the actual Census every 10 years.
This one is supplementary data that helps us, but it is not the
same constitutional mandate, does not have to be done by a
government entity, but is useful to us.
Mr. Goldenkoff. We have not looked into that, no.
Senator Lankford. OK. Well, that may be one that we try to
follow-up on in the days ahead and try to find the most
efficient way to be able to gather this so that we have good
data and information, but we do not lose in cost.
Senator Carper, any additional questions?
Senator Carper. Just one. I do not recall whether you are
married or not, Director Thompson, but every now and then our
spouses have a way of saying to us, ``Well, whatever plans we
had for the rest of our career, you can forget about that.''
Every now and then, my wife--I do not know if it is tongue in
cheek or not, but she will say, ``Somewhere down the line here,
why don't we just put everything we have in storage and just go
and travel around the world.'' And I say, ``OK. When do we have
to do that?'' And so far she keeps letting me off the hook.
But let us say you were married--maybe you are--and you get
into 2016 and your spouse says, ``Everything we have, let us
just put everything in storage and let us spend the next
several years just traveling around the world.'' And, you are
serving this 5-year term. We are grateful that you are doing
it. But in order to maintain peace in the valley with your
spouse, you declined the President's offer for another 5-year
term. And so then we have to figure out how do we get a smooth
handoff, how do we get a smooth handoff carrying up to the next
Census.
One of the reasons we have this 5-year term in law is to
try to ensure we have a better transition. But there is nothing
to ensure that you will want to stay beyond this at this time.
Hopefully you will. But what else are we doing to ensure a
smooth handoff of the Census? I think this is premature, but
what else are we doing and should we be doing to make sure
that, whether you stay or do not, we have that smooth handoff?
Mr. Thompson. Well, thank you, Senator. So there are some
things that are certain, and one is that my term ends December
31, 2016, the term I am currently in. Then there will be
another term, which I could be reappointed for or not. But
there are a lot of things that are going to happen between now
and then, including a Presidential election, that I have no
control over.
So what I am focusing on is putting in place the best plan
for implementing the 2020 Census that I can by the end of my
term so that the Census Bureau has good guidance. That is
something I can control.
I also need to say that the Department of Commerce is being
very supportive of me in that effort. So that is what I am
doing.
Senator Carper. During an earlier administration, I felt we
were at a point where we were really struggling with the
Census. We reached a point where I felt that there was the kind
of engagement from the Secretary of Commerce that we needed.
And you mentioned a couple of times the involvement of the
Secretary and the Secretary's senior staff. Would you just
characterize that again? Because I think it is really
important.
Mr. Thompson. Thank you. The Census Bureau has had
tremendous support from the Office of the Under Secretary for
Economic Affairs, they oversee the Census Bureau and the Bureau
of Economic Analysis. We have had tremendous support from the
Office of the Secretary and from the Secretary herself. I met
with the Secretary periodically, and she wants to make sure
that we are on track to do a good Census and how she can help
me.
And we have had tremendous support from the Office of
Management and Budget, which has understood our needs and has
been very helpful in defining those.
So the Administration is behind conducting an accurate
Census for 2020 and supporting the Bureau in moving forward.
Senator Carper. Good.
Mr. Chairman, it has been an encouraging hearing and an
important one, and I am sorry all of our colleagues could not
be here. On the question of the budget issue that we have
raised and talked a little bit about for 2016, our Chairman is
on the Budget Committee, and he has been out there
participating, as you probably know, in a conference on the
budget. So he is probably in there wrestling and trying to make
sure we get this bump-up so that we will get a good Census in
2020.
Senator Lankford. While we are in here trying to make sure
that we are protected from a $3 billion error.
Senator Carper. There you go.
Senator Lankford. Back and forth again. So thank you all
for being here and what you bring to this. There is a
tremendous gain for the entire Nation. It is a constitutionally
mandated operation, and we are grateful for the service, and
hopefully we can continue to work on efficiencies and bring
down the cost per person and try to examine some efficiencies
out of the box in the days ahead. So thanks. Good hunting on a
lot of the testing and evaluation and things that are happening
to us in the days ahead.
The hearing record will remain open for 15 days until May
the 5 at 5 p.m. for the submission of statements and questions
for the record. Thank you all for being here.
Senator Carper. Mr. Chairman.
Senator Lankford. Yes, sir.
Senator Carper. I want to again mention the name of
Lawrence Buckner who was killed in the line of duty, I think
about 11 days ago.
Senator Lankford. Right.
Senator Carper. I know his name was mentioned earlier.
Senator Lankford. Yes, in the opening.
Senator Carper. You and others extended your condolences to
his family, but sometimes we talk about nameless, faceless
bureaucrats in the Federal Government, and these are people, a
lot of them, who work hard, work long hours, in some cases at
great risk to themselves, and we remember him and thank him and
other of his colleagues who do this dangerous work. Thank you.
Senator Lankford. Entirely appropriate. Thank you again for
the way you are taking care of them and their families.
Thank you. This hearing is adjourned.
[Whereupon, at 4:33 p.m., the Committee was adjourned.]
A P P E N D I X
----------
[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
[all]