[Senate Hearing 111-41]
[From the U.S. Government Publishing Office]


                                                         S. Hrg. 111-41

 
                      LESSONS LEARNED: HOW THE NEW
                 ADMINISTRATION CAN ACHIEVE AN ACCURATE
                     AND COST-EFFECTIVE 2010 CENSUS

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



                                HEARING

                               before the

FEDERAL FINANCIAL MANAGEMENT, GOVERNMENT INFORMATION, FEDERAL SERVICES, 
                AND INTERNATIONAL SECURITY SUBCOMMITTEE

                                 of the

                              COMMITTEE ON
               HOMELAND SECURITY AND GOVERNMENTAL AFFAIRS
                          UNITED STATES SENATE


                                 of the

                     ONE HUNDRED ELEVENTH CONGRESS

                             FIRST SESSION

                               __________

                             MARCH 5, 2009

                               __________

       Available via http://www.gpoaccess.gov/congress/index.html


                       Printed for the use of the
        Committee on Homeland Security and Governmental Affairs



                  U.S. GOVERNMENT PRINTING OFFICE
49-639                    WASHINGTON : 2009
-----------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Printing 
Office  Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800; 
DC area (202) 512-1800 Fax: (202) 512-2104  Mail: Stop IDCC, 
Washington, DC 20402-0001




        COMMITTEE ON HOMELAND SECURITY AND GOVERNMENTAL AFFAIRS

               JOSEPH I. LIEBERMAN, Connecticut, Chairman
CARL LEVIN, Michigan                 SUSAN M. COLLINS, Maine
DANIEL K. AKAKA, Hawaii              TOM COBURN, Oklahoma
THOMAS R. CARPER, Delaware           JOHN McCAIN, Arizona
MARK PRYOR, Arkansas                 GEORGE V. VOINOVICH, Ohio
MARY L. LANDRIEU, Louisiana          JOHN ENSIGN, Nevada
CLAIRE McCASKILL, Missouri           LINDSEY GRAHAM, South Carolina
JON TESTER, Montana
ROLAND W. BURRIS, Illinois
MICHAEL F. BENNET, Colorado

                  Michael L. Alexander, Staff Director
     Brandon L. Milhorn, Minority Staff Director and Chief Counsel
                  Trina Driessnack Tyrer, Chief Clerk
                                 ------                                

 SUBCOMMITTEE ON FEDERAL FINANCIAL MANAGEMENT, GOVERNMENT INFORMATION, 
              FEDERAL SERVICES, AND INTERNATIONAL SECURITY

                  THOMAS R. CARPER, Delaware, Chairman
CARL LEVIN, Michigan                 JOHN McCAIN, Arizona
DANIEL K. AKAKA, Hawaii              TOM COBURN, Oklahoma
MARK L. PRYOR, Arkansas              GEORGE V. VOINOVICH, Ohio
CLAIRE McCASKILL, Missouri           JOHN ENSIGN, Nevada
ROLAND W. BURRIS, Illinois

                    John Kilvington, Staff Director
    Bryan Parker, Staff Director and General Counsel to the Minority
                       Monisha Smith, Chief Clerk



                            C O N T E N T S

                                 ------                                
Opening statements:
                                                                   Page
    Senator Carper...............................................     1
    Senator McCain...............................................     3
    Senator Burris...............................................     4
    Senator Coburn...............................................     5

                               WITNESSES
                        Thursday, March 5, 2009

Barbara Everitt Bryant, Ph.D., Former Director, U.S. Census 
  Bureau.........................................................     6
John Thompson, President, National Opinion Research Council......     8
Robert Goldenkoff, Director, Strategic Issues, U.S. Government 
  Accountability Office..........................................    10
David Powner, Director, Information Technology Management Issues, 
  U.S. Government Accountability Office..........................    12
Lawrence B. Brown, Ph.D., Chair, Committee on National 
  Statistics, National Academy of Sciences.......................    13
Robert B. Hill, Ph.D., Sociologist and Former Chair of the 
  Advisory Committee on African American Population, U.S. Census 
  Bureau.........................................................    14

                     Alphabetical List of Witnesses

Brown, Lawrence B., Ph.D.:
    Testimony....................................................    13
    Prepared statement...........................................    78
Bryant, Barbara Everitt, Ph.D.:
    Testimony....................................................     6
    Prepared statement...........................................    38
Goldenkoff, Robert:
    Testimony....................................................    10
    Prepared statement...........................................    49
Hill, Robert B., Ph.D.:
    Testimony....................................................    14
    Prepared statement...........................................    95
Powner, David:
    Testimony....................................................    12
    Prepared statement...........................................    68
Thompson, John:
    Testimony....................................................     8
    Prepared statement...........................................    44

                                APPENDIX

Hon. Vincent P. Barabba, Former Director of the U.S. Census 
  Bureau, prepared statement.....................................    33
Letter to Thomas L. Mesenbourg, Acting Director and Deputy 
  Director, U.S. Census Bureau, dated February 19, 2009, from Mr. 
  Brown..........................................................    82


                      LESSONS LEARNED: HOW THE NEW



                 ADMINISTRATION CAN ACHIEVE AN ACCURATE



                     AND COST-EFFECTIVE 2010 CENSUS

                              ----------                              


                        THURSDAY, MARCH 5, 2009

                                 U.S. Senate,      
        Subcommittee on Federal Financial Management,      
               Government Information, Federal Service,    
                              and International Security,  
                          of the Committee on Homeland Security    
                                        and Governmental Affairs,  
                                                    Washington, DC.
    The Subcommittee met, pursuant to notice, at 2:31 p.m., in 
room SD-342, Dirksen Senate Office Building, Hon. Thomas R. 
Carper, Chairman of the Subcommittee, presiding.
    Present: Senators Carper, Burris, McCain, and Coburn.

              OPENING STATEMENT OF SENATOR CARPER

    Senator Carper. The Subcommittee will come to order. We 
welcome you, one and all. I am delighted that Senator McCain is 
able to join us today, and to our guests, as well.
    The Subcommittee has responsibility for oversight in a 
number of areas. One of those is with respect to the Census 
Bureau's preparation for the 2010 Census, and today we are 
going to hear from former Census Directors and experts within 
the statistical community who will offer their valuable 
insights into lessons learned from past Censuses. It is my hope 
that their experiences can help the Census Bureau conduct an 
accurate and cost-effective Census in 2010.
    There is a well-known adage that knowledge is power and the 
Census is an important source of knowledge and information. 
Census data empowers citizens at every level of government and 
are integral in achieving equitable political representation 
and fair allocation of resources. Finding and enumerating 
nearly 300 million individuals in the correct location is, of 
course, an extremely daunting task. The 2000 Census involved 
the hiring of nearly half-a-million temporary workers, the 
opening of some 500 local Census offices nationwide, 
processing, I believe, 1.5 billion sheets of paper, and 
following up with 42 million non-responsive households.
    Given the sheer magnitude of such an undertaking, a 
shortcoming in one area can quickly have a domino effect on 
other operations. For example, a low mail response rate would 
increase the non-response follow-up workload, which in turn 
would drive the Bureau's staffing needs and drive up costs.
    With each Census, the challenge continues to grow in terms 
of cost and complexity as our population becomes larger, more 
diverse, and increasingly difficult to enumerate. The cost of 
the 2010 Census has escalated to an estimated $14 billion, and 
that is what my statement says. The cost of the 2010 Census has 
escalated to an estimated $14 billion, and I think that is 
true, making it the most expensive in the history of our 
country. Put another way, it will cost the Nation an estimated 
$100 or so to count each household in 2010, compared with about 
$56 in 2013 dollars in 1970. The growing cost of the Census at 
a time when the Federal Government is facing unprecedented 
budget deficit highlights the importance of making sure that 
every additional dollar spent on the Census actually improves 
the quality of the data.
    Although the 2000 Census was an improvement when compared 
to the 1990 Census, there were still many deficiencies. In 
2000, 6.4 million people were missed and 3.1 million people 
were counted twice, producing a net undercount of some 3.3 
million people.
    I just interject, usually when we have an overcount, it is 
people that have more than one house, maybe a second home or a 
vacation home, or maybe they have a child who is in college in 
another State, and those are situations that lead to 
overcounts, and they usually occur among the more affluent 
families. On the other hand, the undercounts usually occur 
among a lot of minority families, whether African American or 
Latino or Native American. But we ended up with an undercount 
of about 6 million people, for the most part lower-income 
folks, and an overcount of about 3 million people the last time 
we did this, mostly of more affluent people. Neither one is a 
good situation, but that is not what we need for this Census 
and the conducting of this Census.
    At any rate, the 2010 Census is approaching rapidly, as we 
know, with the Census date less than 13 months away. The Bureau 
has faced many operational and organizational challenges that 
have jeopardized its success. These challenges include 
underfunding for outreach to minority communities and the 
colossal mismanagement and failures of the contract for hand-
held computers that led to an entire replan of the Census very 
late in the game. Senator Coburn and I have been working on 
this for several years under his leadership as the Chairman of 
this Subcommittee and more recently under my own.
    Further, I understand that the Bureau lacks plans for 
testing some of its key information technology systems. With 
such a substantial reliance on new technology, a robust testing 
strategy is necessary to identify and correct any problems that 
may arise.
    I believe we are at a critical juncture. I don't think it 
is overstating things to say that the 2010 Census is 
approaching a state of emergency. Significant work still has to 
be done, and the Bureau does not have a Director in place to 
assist them in making these critical decisions. Last month, I 
sent a letter to President Obama urging him to nominate a new 
Director as soon as possible. I renewed that request as 
recently as this morning.
    It is my hope that we have learned from these valuable 
lessons and can continue to work together to ensure the success 
of the 2010 Census.
    We look forward to the expert testimony here today from our 
distinguished panel of witnesses.
    I would just say to our colleagues, and we have been joined 
by Senator Coburn, who knows these issues as well or better 
than me--Senator Coburn said that he didn't think the people 
here knew who Senator McCain was so he is helping with his name 
tag. [Laughter.]
    But, Senator Coburn, we are in a situation where we have 
gone from an Administration where we didn't have a Census 
Director in place for the longest time, we had to wait for a 
long time to get a nominee, finally got a nominee, a very good 
one, and he stayed with us for about a year and the beginning 
of this year he tendered his resignation letter along with a 
whole lot of other appointed officials. So now we are waiting 
until we get a good solid replacement and time is wasting.
    Senator McCain, it is great to have you here.

              OPENING STATEMENT OF SENATOR MCCAIN

    Senator McCain. Thank you very much, Mr. Chairman. I want 
to thank you for holding this hearing and I want to thank the 
witnesses for appearing here today. I again thank you and 
Senator Coburn, who has a great sense of humor, I am sure you 
will note, for your attention to this issue.
    The first thing I would like to point out, I think it is 
pretty obvious that we need to have a Director of the Census. I 
am sure our witnesses will be in total agreement on that aspect 
of this issue.
    We are nearing the time where certainly final preparations 
for one of the more important events are underway--I don't 
think people appreciate the importance of the Census. It not 
only means a lot to us as far as Congressional districts are 
concerned, but so much of our Nation's operations as far as 
apportionment of money for various programs, apportionment of 
responsibilities. So much of the things that we do, and, in 
fact, laws that we pass every day are implemented through the 
Census, guided by our knowledge, hopefully knowledge or lack of 
knowledge of the people we have in each State and each part of 
each State.
    So obviously a fair and accurate counting is critical. No 
State should be unfairly denied representation or funding for 
essential services because the Census Bureau can't resolve 
problems that have plagued us for decades, and obviously one of 
those problems is undercounting. I guarantee you that whether 
it is totally accurate or not, there will be States that claim 
undercounting at the end of this process. I think our witnesses 
would agree with that.
    And I am one of them. My home State of Arizona suffered 
severely from undercounting in the 1990 Census. As a result, 
Arizona was denied an additional Congressional seat and lost 
millions of dollars in Federal revenue for schools, roads, 
housing, and other public services.
    Resolving the issue continues to spur debate and the need 
to make constant improvements to traditional enumeration 
methods remains a top priority. So the process must be fair. It 
has got to be conducted in a manner that doesn't discriminate 
and doesn't dissuade participation.
    I want to point out again that some recent activity on the 
part of the Census Bureau does not lend itself to increasing 
the confidence level. Of course, I am talking about the 
investment of millions of dollars in hand-held computers that 
can't deliver the capability that was once promised. I know our 
Subcommittee examined the issue, but I still feel compelled to 
bring this up again because it is a terrible precedent to set 
and does not give us confidence. So it lost taxpayers' money 
because of cost overruns and lost productivity. We can't 
afford, obviously, to waste that.
    I am concerned that we have enough time, and I will be 
interested in hearing from our witnesses, about adequate 
testing of all critical systems and procedures before 
additional Census activities begin. I am most interested to 
hear from our GAO witnesses about the current status of these 
setbacks and how much delay this mismanagement has caused.
    I want to thank you, Mr. Chairman. I want to thank our 
witnesses. Again, this is a very important process that this 
Nation is about to embark on. You mentioned people that have 
second homes. I also think that this is a period of great 
mobility in America from one place to another for economic 
reasons and others, including the traditional mobility of 
Americans. I do not know, maybe one of our witnesses knows how 
many people move from one State to another in the course of a 
year, but it is significant and on the increase.
    We have an obligation to ensure that every American is 
counted and counted accurately and I thank you, Mr. Chairman, 
and I want to thank Senator Coburn for his involvement for many 
years in this issue. Thank you.
    Senator Carper. Thank you for your statement and thank you 
for joining us in this effort.
    I am going to go to Senator Burris and then to Senator 
Coburn.

              OPENING STATEMENT OF SENATOR BURRIS

    Senator Burris. Thank you very much, Mr. Chairman.
    Senator Carper. Welcome. We are delighted that you are 
here.
    Senator Burris. My pleasure. I would like to extend my warm 
welcome to our distinguished valued panelists. I am so pleased 
to see that you are carrying out your constitutional and 
committed responsibilities to inform us as officials.
    As we approach what projections indicate will be the most 
expensive Census in history, we must assure that sound 
leadership aids its execution. It is crucial that we take 
action quickly to guarantee success. Constituents must trust us 
to spend their money wisely and we must ensure that we plan 
both comprehensive and responsible.
    America has changed greatly in the last 10 years and I fear 
that some citizens may be less likely to participate in the 
Census. We already face significant language barriers and we 
now must attempt to overcome suspicion of legitimacy. With 
identity theft on the rise, many ordinary Americans may 
overlook the necessity of the Census for fear of their personal 
safety, so we must be very concerned about what is happening in 
the minds of our citizens. Fortunately, through the knowledge 
of the previous experience, hard work, and development of new 
technology, significant opportunities now exist for us to 
improve the process, and I am with Senator McCain on his 
comments.
    The undercount, we hear so much about the undercount. I 
live in Chicago and in a metropolitan area, we are always 
complaining about the fact that in those areas, we are not 
counted because nobody wants to go up into especially what we 
call the developments. Some people refer to them as projects. A 
lot of them have been torn down in Chicago now, but we hope to 
find where all those people were relocated and try to identify 
those people because that is where a lot of the allocations are 
determined.
    I will have some questions later, and Mr. Chairman, I might 
have to run out, but I will come back with some questions later 
on.
    Thank you, Mr. Chairman.
    Senator Carper. Senator Burris, thank you very much.
    I am delighted again that my partner in this initiative and 
part of my good oversight is here with us, Senator Coburn. 
Please proceed.

              OPENING STATEMENT OF SENATOR COBURN

    Senator Coburn. I will just spend a short amount of time. 
The best recommendation I could make to the President is to 
bring Mr. Murdock back. He did a great job while he was there. 
I think it is unfortunate for us as a Nation that he left in 
the midst of straightening out a lot of the problems that were 
there. I will save all the rest of my comments for the time of 
which we have questioning.
    Senator Carper. All right. Thanks very much.
    Let me briefly introduce our witnesses, and we will start 
with the Hon. Barbara Bryant. Welcome. She was Director of the 
Census Bureau from 1983 to 1991, and during her tenure, she 
directed the 1990 Decennial Census of Population and Housing, 
the 1992 Economic and Agricultural Census, and other major 
surveys. I believe you were appointed by President George 
Herbert Walker Bush and confirmed by the Senate as the first 
woman to head the Census Bureau in 200 years of Census taking. 
Dr. Bryant received her Bachelor's degree from Cornell 
University and her M.A. and Ph.D. from Michigan State 
University--a Spartan. Welcome today.
    John Thompson, I call him the real John Thompson, is the 
President of the National Opinion Research Council at the 
University of Chicago. Mr. Thompson came to the Council after a 
27-year career at the Census Bureau, where as one of the 
Bureau's most senior career officers he had the responsibility 
for all aspects of the 2000 Census, including management, 
operations, and methodology. Mr. Thompson attended Virginia 
Tech University, where he was awarded a B.S. and a Master's of 
Science degree in mathematics.
    Next, Robert Goldenkoff, Director of Strategic Issues at 
GAO, where he is responsible for reviewing the 2010 Census and 
government-wide human capital reforms. Mr. Goldenkoff has also 
performed research on issues involving transportation security, 
human trafficking, and Federal statistical programs. He 
received his Bachelor's in political science and Master's in 
public policy from George Washington University.
    Dave Powner, good to see you again. Thank you for joining 
us. He has over 20 years of experience in information 
technology issues in both the public and private sector. He is 
currently responsible for a large segment of GAO's information 
technology work, including systems development, IT investment 
and management, health IT, and cyber critical infrastructure 
protection reviews. He is no stranger to this Subcommittee. 
Thank you for joining us again today.
    Lawrence D. Brown, Professor in the Department of 
Statistics at the Wharton School at the University of 
Pennsylvania. He is a member of the National Academy of 
Sciences and has served on several committees and panels of the 
National Research Council, including the Committee on National 
Statistics. Dr. Brown's work includes evaluations on the design 
and methodology of both the 2000 and the 2010 Census programs. 
He received a Bachelor of Science degree in mathematics from 
the California Institute of Technology and a Ph.D. in math 
statistics from Cornell.
    And last but not least, Robert Hill, a sociologist who 
recently retired as Senior Researcher at Westat, a research 
firm in Rockville, Maryland. He was Chair of the U.S. Census 
Bureau Advisory Committee on the African American Population 
for both the 1980 and, I think, the 2000 Censuses. Dr. Hill 
received his Bachelor's of Art in sociology from the City 
College of New York and a Doctorate in sociology from Columbia 
University.
    We would have one other person here with us today, Vincent 
Barabba, former Census Director during the Nixon and Carter 
administrations. He was scheduled to participate in our hearing 
today. Due to a series of unanticipated events, he is unable to 
join us. His testimony will be submitted for the record and the 
Subcommittee looks forward to working with him in the future as 
we continue our oversight of the 2010 Census.\1\
---------------------------------------------------------------------------
    \1\ The prepared statement of Mr. Barabba appears in the Appendix 
on page 33.
---------------------------------------------------------------------------
    With those introductions behind us, let me just say, Ms. 
Bryant, we welcome you here. We are delighted that you are 
going to be our lead-off hitter. We will go right down the line 
and then we will ask questions. Thank you.

TESTIMONY OF BARBARA EVERITT BRYANT, PH.D.,\2\ FORMER DIRECTOR, 
                       U.S. CENSUS BUREAU

    Ms. Bryant. Thank you, Chairman Carper and Acting Ranking 
Member McCain. You have been given my introduction statement.
---------------------------------------------------------------------------
    \2\ The prepared statement of Ms. Bryant appears in the Appendix on 
page 38.
---------------------------------------------------------------------------
    Lesson one for the 2010 Census is the one that all of you 
have referred to, and that is the importance of getting a new 
Census Director in immediately. I am pleased to see you are 
pressing on the President to make the nomination. On this, I 
really speak from experience, because----
    Senator Carper. If I could interrupt, I also pressed this 
morning on the President's nominee for Commerce. I ran into 
Governor Locke yesterday and again today and I said, if you 
haven't started thinking about who you would like to have on 
the Census Bureau, start thinking about it right now.
    Ms. Bryant. Thank you very much----
    Senator Carper. I brought it to his attention.
    Ms. Bryant [continuing]. From all of us.
    Twenty years ago, I became Director in the same election 
cycle we are in now, that is in the Presidential election in 
the year ending in eight and nomination in the year ending in 
nine. However, I was not in office until December 7, 1989, 3 
weeks before the start of the Census year. I was eventually 
confirmed by the Senate.
    Would Census procedures have been different if I had been 
in office sooner? Definitely, and I elaborate on this in my 
written testimony.
    I am a supporter of making the job of Director of the 
Census Bureau a 5-year Presidential appointment, starting in 
the years one and six. The planning cycle for operations as 
large as the Decennial and Economic Censuses are long and only 
with a several-year lead time can a Director have any real 
input into what is going to be done. I also will say it is no 
fun for a Census Director to sit in front of Congressional 
committees like this defending operations in which he or she 
had no input.
    Lesson two is that a major professional coordinated 
communications and advertising campaign is vital to Census 
success and accuracy. Such a campaign requires a major 
financial outlay. It has two components, a large volume of 
inexpensive promotional materials that can be handed out or 
posted at the local level; and radio, TV, and newspaper spots 
professionally produced with goals of reaching both the mass 
national and targeted audiences.
    The advertising campaign has got to be on a scale 
comparable to what a private sector firm would use to introduce 
a new product. After all, the Census is a new product to 
everyone in their 20s, and is a 10-year-old, half-forgotten 
product for anyone 30 and over.
    Lesson three is outreach to hard-to-count segments of the 
population. Through partnerships with geographic, ethnic, and 
racial organizations, we can help reduce the undercount. The 
Census historically has fully counted some segments of the 
population, particularly homeowners and older Americans. It 
falls short of fully counting the very mobile, the renters, the 
young people, and particularly those in Hispanic, Latino, 
African American, and American Indian communities. These hard-
to-count are best reached with one-on-one contacts from local 
people and organizations in whom they have trust.
    Communicating the fact that the Census Bureau will not give 
information from their Census forms to any other organization 
or individual is a very hard message to get across. Only 
trusted sources can convince the reluctant, the fearful, or the 
uninformed that the Census Bureau does not give information to 
the INS, the IRS, landlords, ex-spouses, or mothers-in-law. 
[Laughter.]
    In addition to implementing these three lessons, three 
other factors will help improve the accuracy in 2010. One is 
the American Community Survey, which is now ongoing, and it 
replaces the information formerly gathered on the long form 
with about 50 questions that went to 17 percent of households. 
The long form always had a several percent lower mail return 
than the short form.
    The second thing that is going to help is the downturn in 
employment. This is a national tragedy, but it does have the 
benefit for the Census Bureau that they are going to have a 
bigger pool of people from which to choose their temporary 
employees. My Census, we had 5 percent unemployment and the 
pool had dried up to be a puddle.
    And third, the undercount research done after all the 
recent Censuses identifies very precisely where non-respondents 
are geographically so they can be targeted. Such research is 
important to fund for every Census.
    But now the big inhibitor to a good count in 2010 is the 
fear in the Hispanic-Latino communities. The current 
Immigration and Naturalization raids on such communities, on 
employers and neighborhoods is bound to depress cooperation. I 
flew in yesterday from Phoenix, Senator McCain, where I do have 
a second home and some households there include both legal and 
undocumented immigrants. You can imagine when a Census taker 
goes into those neighborhoods and says, ``I am from the Census 
Bureau and I have a few questions from the Government,'' what a 
warm reception they may receive.
    Finally, you asked us to comment on a cost-effective 
Census. Well, counting every person and household is never 
going to be a cheap operation, particularly with all this 
follow-up on the hard-to-count. But with this experience in the 
logistics of the operation and its magnitude, and I will add 
the wonderful employees, the long-term career employees at the 
Census Bureau, the Census Bureau probably does as cost-
effective a job as any organization could. But current staff, 
not the Director of a $2.6 billion Census conducted 20 years 
ago must report to you on present efforts to be cost effective.
    Senator Carper. Thank you very much. Mr. Thompson.

  TESTIMONY OF JOHN THOMPSON,\1\ PRESIDENT, NATIONAL OPINION 
                        RESEARCH COUNCIL

    Mr. Thompson. Thank you, Mr. Chairman and Members of the 
Subcommittee. I am truly honored to be glad to speak to you 
about the 2010 Census. As you asked, I will talk about some of 
the successes in 2000 and relate them to the risks with respect 
to the 2010 Census.
---------------------------------------------------------------------------
    \1\ The prepared statement of Mr. Thompson appears in the Appendix 
on page 44.
---------------------------------------------------------------------------
    Before I start, I would quickly like to recognize the fact 
that I worked with a lot of the people at the Census Bureau 
right now. I know that they are motivated to do high-quality 
work. They are nonpartisan and they are very good and my 
remarks are intended to help them with their effort, not in any 
way to criticize them.
    So starting with Census 2000, I think the first factor I 
would mention in success, and you will hear a lot of the same 
things, I think, is the unprecedented support that the 2000 
Census received. One example of that was that the mail response 
rate was 67 percent, which was higher than the 1990 Census rate 
of 65 percent. That was the first Census where the decline in 
mail response rate had been reversed.
    I think there were three factors that contributed to that. 
The first is a paid advertising campaign. It was the first 
Census that used a paid advertising campaign.
    The second was a very well-funded effort to establish 
partnerships with local community groups.
    And the third was a very effective communications strategy 
that reached out to numerous stakeholders, including the 
Congress, State, local, and Tribal governments, and a variety 
of advocacy groups.
    In addition, the non-response follow-up operation, the most 
difficult part of the Census, was finished in 9 weeks. One of 
the factors that contributed to that was, in my opinion, the 
fact that the public was highly motivated to cooperate and we 
didn't have to visit the households too many times. There were 
other factors which I will mention, too.
    Another factor in the success of the Census was that we had 
a Director in place in time to provide leadership and guidance. 
That Director was Kenneth Pruitt. I had the privilege of 
working with him. He provided leadership, guidance, and set an 
environment up where the career people could be successful. I 
also would note that I had the pleasure of working with Dr. 
Bryant on the 1990 Census and that was also a fine experience.
    Another factor was that we had a very strong management 
team in place that was very experienced and included managers 
with both Census experience and managers from outside of the 
Census Bureau that brought different perspectives to problem 
solving, and we had the team in place in time to make some 
significant contributions.
    Another factor was that the Census 2000 field effort was 
very well funded and well managed. We had done studies that 
linked pay for retention and we had an excellent group of 
regional directors who managed the Census. We were able to 
recruit and retain a workforce to do the job.
    We also had effective usage of private sector contractors 
in 2000 to provide advance technology solutions to our data 
capture operations. We had private sector contractors that ran 
some very large facilities, recruited a lot of staff, and put 
in place optical scanning and intelligent character recognition 
software that allowed us to capture over 80 percent of the 
handwritten entries on the Census forms with a very high degree 
of accuracy.
    The final factor I will mention which is relevant to this 
Census is that we had a thorough testing of all of our 
operational systems. We had a dress rehearsal in 1998 where we 
tested our systems from start to finish. That was very 
important. We also, I will note, had to change the Census 
design. There was a controversy over Census. We were going to 
add two tracts, and in 1999, we decided--the Supreme Court 
decided that we would not use sampling for the count, so we had 
a redesigned Census that did not use sampling. The fact that we 
had our systems tested, we were able to modify them and move 
forward.
    So for 2010, the major risk is in systems development and 
testing. I think the Subcommittee said that and it is fairly 
obvious. They had to abandon their plan to use hand-helds. They 
are going back to a paper-based system. They have taken over 
the control system from a private contractor. I think the best 
recommendation I can make is that they need to do a full-scale, 
large field test of that system to conduct a response follow-
up, to collect some data, to make sure that they have 
interfaces with all the key systems, and to make sure that 
interviewers, Census enumerators, fairly inexperienced people 
can utilize these systems.
    I will say just a couple more things. We have talked about 
the Director. I agree with that.
    Senator Carper. When you say agree with it, be more 
specific in what you agree with, the importance of getting a 
good one or----
    Mr. Thompson. I agree that we need a Director of the Census 
Bureau as soon as possible. Like I said, having a Director in 
place during the Census is just incredibly important, and in 
the period preceding the Census.
    They need also to establish a communications lead. That 
needs to be appointed. They don't have one right now.
    The management staff, I think they are doing a good job. 
They are working as hard as they can. I think they are very 
thin. I think they should reach out to some of the other 
Federal agencies to look for some more talent.
    Their coverage measurement system is currently scheduled in 
a fashion where I am concerned that it won't produce accurate 
measures. It is scheduled to take place too far after Census 
Day and I think there will be issues with recall bias 
associated with it. They need to tighten the schedule up 
similar to previous Censuses.
    And finally, Dr. Brown will talk a lot, I think, about the 
experimental program for the 2020 Census, but I encourage the 
Subcommittee to challenge the Census Bureau to develop plans 
for a different type of Census for 2020, one that is less 
expensive and relies on alternative methods of data collection.
    That concludes my remarks. Thank you.
    Senator Carper. Thank you, Mr. Thompson. Mr. Goldenkoff.

TESTIMONY OF ROBERT GOLDENKOFF,\1\ DIRECTOR, STRATEGIC ISSUES, 
             U.S. GOVERNMENT ACCOUNTABILITY OFFICE

    Mr. Goldenkoff. Chairman Carper, Ranking Member McCain, 
Members of the Subcommittee, thank you for the opportunity to 
be here today to provide a progress report on the 2010 Census. 
I am here with Dave Powner, a Director on GAO's Information 
Technology team.
---------------------------------------------------------------------------
    \1\ The prepared statement of Mr. Goldenkoff appears in the 
Appendix on page 49.
---------------------------------------------------------------------------
    As requested, in our remarks today, I will provide a broad 
overview of the status of key Census-taking operations and Mr. 
Powner will focus on the findings and recommendations contained 
in our report on IT testing which we are releasing today.
    This afternoon's hearing is particularly timely. It was 
exactly one year ago today that GAO designated the 2010 Census 
as a high-risk area for three reasons. First, there were 
weaknesses in the Census Bureau's IT acquisition and contract 
management function. Second, there were problems with the 
performance of hand-held computers used to collect data. And 
third, the ultimate cost of the Census is uncertain, although 
it is currently estimated at more than $14 billion.
    At the same time, just over one year from now, it will be 
Census Day. Little time remains to address the challenges that 
have emerged thus far and make final preparations for the 
numerous operations that will take place throughout 2010. In 
short, today's hearing is a convenient waystation on the road 
to Census Day, a time to look back on the Census Bureau's 
efforts over the past year to address the operational 
challenges that have emerged thus far as well as to look ahead 
to what the Bureau needs to do in the coming months to help 
ensure a successful head count.
    Importantly, the Bureau has made commendable progress over 
the past year in rolling out key components of the Census and 
to strengthen certain risk management efforts. Still, the 
Census remains high-risk because a dress rehearsal of all 
Census operations that was planned for 2008 was curtailed. As a 
result, critical activities, including some that will be used 
for the first time in a Census, were not tested in concert with 
one another or under Census-like conditions.
    The bottom line is that key Census-taking activities, 
including those that will ultimately drive the final cost and 
accuracy of the count, continue to face challenges and the 
Bureau's overall readiness for 2010 is uncertain.
    One such challenge is building the Bureau's address list. 
Because a complete and accurate address list is the cornerstone 
of a successful Census, the Bureau has a number of operations 
aimed at including every residence in the country and works 
with the U.S. Postal Service, agencies at all levels of 
government, as well as a number of non-governmental entities. 
In a few weeks, the Bureau will send thousands of workers to 
walk every street in the country to update the Census address 
list and maps in an operation called address canvassing. Census 
workers will use hand-held computers to collect data.
    As you know, when the devices were tested, they experienced 
performance problems, such as freeze-ups and unreliable 
transmissions. The Bureau took steps to fix these issues and 
the results of a small-scale test held last December are 
encouraging. Nonetheless, more information is needed to 
determine the Bureau's overall readiness for address 
canvassing, as a field test was not an end-to-end systems 
evaluation, did not validate training, help desk support, and 
other requirements, and did not include urban areas.
    Uncertainties also surround the Bureau's ability to 
implement operations that will be used for the first time in a 
decennial Census, including the targeted second mailing to 
reduce the non-response follow-up workload and the need to 
fingerprint temporary Census workers. The Bureau's readiness 
for these activities is uncertain because they have not been 
tested under Census-like conditions.
    Another challenge facing the Bureau, as we have mentioned, 
is reducing the undercount. As with past enumerations, the 
Bureau is putting forth tremendous effort to reach groups that 
are often missed by the Census, such as minorities, renters, 
and people with limited English proficiency. For example, the 
Bureau plans to provide language assistance guides in 59 
languages, an increase from 49 languages in 2000.
    Although the effects of the Bureau's communication efforts 
are difficult to measure, the Bureau reported some positive 
results from its 2000 marketing efforts with respect to raising 
awareness of the Census. Still, a longstanding challenge for 
the Bureau is converting awareness of the Census into an actual 
response.
    Some specific hurdles that need to be overcome include the 
Nation's linguistic diversity and privacy concerns and a post-
September 11, 2001 environment that could heighten some groups' 
fears of government agencies.
    In summary, just 13 months remain until Census Day. At a 
time when major testing should be completed and there should be 
confidence in the functionality of key operations, the Bureau 
instead finds itself managing late design changes and 
developing testing plans. The Bureau has taken some important 
steps towards mitigating some of the challenges that it has 
faced to date, yet much remains uncertain and the risks to a 
successful decennial Census continue.

TESTIMONY OF DAVID POWNER,\1\ DIRECTOR, INFORMATION TECHNOLOGY 
    MANAGEMENT ISSUES, U.S. GOVERNMENT ACCOUNTABILITY OFFICE

    Mr. Powner. Chairman Carper, Ranking Member McCain, and 
Members of the Subcommittee, the accuracy of the 2010 Census 
depends in large part on the proper functioning of IT systems 
both individually and when integrated together.
---------------------------------------------------------------------------
    \1\ The prepared statement of Mr. Powner appears in the Appendix on 
page 68.
---------------------------------------------------------------------------
    Mr. Chairman and Dr. Coburn, your oversight of the Bureau's 
acquisition of IT systems was critical last year. In 
particular, the field data collection system is no longer 
spiraling out of control and that contract is $500 million less 
than the initial estimates provided at your hearings last 
summer. Your oversight is needed once again in the technology 
area to ensure that between now and Census Day, these systems 
are rigorously tested.
    Today, we are releasing our latest report completed at your 
request which highlights that significant testing remains. Six 
major systems need to complete system testing and much 
integration testing needs to occur. Plans for conducting this 
testing are not completely in place. In order to ensure 
effective test execution, the Bureau needs comprehensive 
metrics to monitor test completion and effective executive-
level oversight to keep the pressure on and to manage risks.
    Our report contains 10 detailed recommendations that the 
Bureau has agreed to address. For example, integration testing 
includes testing the interfaces or the handshake between 
systems. Our work found that not only are there not complete 
plans for integration testing of these interfaces, but there is 
not even a master list of interfaces. Not having such basic 
information at this stage is unacceptable and our 
recommendations call for the Bureau to, one, develop a master 
list of interfaces; two, prioritize the interfaces based on 
criticality and need date; and three, to use this information 
to develop all the needed integration test plans.
    To the Bureau's credit, we are seeing more plans and better 
metrics, but there is still much work ahead in both areas. I 
would like to stress the need to prioritize. It is likely that 
the Bureau will not have enough time to test everything. 
Testing the most important aspects of certain systems, 
interfaces, and operations is critical given the limited time 
remaining.
    Mr. Chairman, again, thank you for your leadership and I 
will look forward to your questions.
    Senator Carper. Thank you very much, Mr. Powner, for you 
and others of your colleagues at GAO for helping Senator Coburn 
and I and our staffs in this effort. Thank you.
    Dr. Brown, please.

 TESTIMONY OF LAWRENCE D. BROWN, PH.D.,\1\ CHAIR, COMMITTEE ON 
       NATIONAL STATISTICS, NATIONAL ACADEMY OF SCIENCES

    Mr. Brown. Good afternoon, Mr. Chairman. Thank you for 
inviting me to talk to you this afternoon. As you noted, I am a 
professor of statistics at the Wharton School of Business and I 
have been actively interested in issues relating to the 
Decennial Census for over a decade.
---------------------------------------------------------------------------
    \1\ The prepared statement of Mr. Brown appears in the Appendix on 
page 78.
---------------------------------------------------------------------------
    Among other things, I have served on several other National 
Academy of Sciences advisory panels involving Census issues, 
and currently I am Chair of a panel to review the Census 
program of evaluations and experiments. Many of my comments 
this afternoon are drawn from a very recent letter report of 
this panel that was mailed to Thomas Mesenbourg as Acting 
Director of the Census Bureau.\2\
---------------------------------------------------------------------------
    \2\ The letter from Mr. Brown appears in the Appendix on page 82.
---------------------------------------------------------------------------
    There are three issues I would like to bring to your 
attention from our panel's reports. Two of these involve 
research and planning that should be part of the 2010 Census 
and the third concern is a more immediate one about research 
that should be conducted before fielding the 2010 Census, and I 
will start with that issue.
    Actually, I am the third person on this panel, as well as 
you, to have talked about the concern with the operating 
control system. As you have remarked and several others have 
remarked, the hand-held devices that were scheduled to be used 
in the non-response follow-up portion of the Census, 
academically termed NRFU, were withdrawn from use and that led 
to a considerable increase in Census costs.
    But I want to focus on a different aspect of this forced 
change and it is really the aspect that both John Thompson and 
Robert Goldenkoff have also mentioned. These devices had been 
designated to form the core of the operating control system for 
NRFU. NRFU, as you noted, is a process that has over half-a-
million people in the field operating out of many local 
offices. This army of people requires a system to keep track of 
it and the Census Bureau is now in the process of restructuring 
their entire operating control system because of the necessity 
of removing the hand-helds.
    So because of the timing of the decision to revert from 
hand-held computers to paper-based NRFU, the 2008 dress 
rehearsal did not test NRFU at all, and this was, of course, a 
major gap in Census testing. Because it wasn't tested, the 
dress rehearsal provided no information on interaction of NRFU 
processes with the redesigned coverage follow-up operation and 
various other components of the process.
    As a remedy, the Census Bureau has scheduled a number of 
isolated component tests, but this testing, component-wise 
testing strategy puts the Bureau in an extremely risky 
position. So I want to just reemphasize that the Bureau needs 
to perform as full and realistic an operational test from start 
to finish of this system as they can, including all of the 
interactions among the various components.
    So the two research issues that I want to mention, and I 
will try and be brief in mentioning them, the first of these 
involves administrative records that could be used in the 
Census. The Census Bureau in the past two decades has conducted 
a research program to see whether administrative records could 
be used to increase the accuracy and reduce the cost of the 
Census. We believe that they offer the best chance of 
accomplishing those ends, but there are not scheduled to be any 
major tests of administrative records in the 2010 Census, and 
given that their use provides one of the few opportunities to 
substantially reduce Census field costs, we believe that the 
Census Bureau should devote serious effort and attention to 
including an experiment or research during the 2010 Census to 
see whether such records can be used in the future.
    And finally, with respect to the Internet, the Internet is 
another opportunity for cost reduction and improvement in data 
quality and the Census Bureau has no plans to incorporate 
Internet questionnaires in 2010 or to perform research on how 
that would enable them to be used in 2020. So we believe that, 
if for no other reason than to avoid looking out of step with 
modern data collection and because of their problems, the 
Bureau should be conducting some coordinated program of 
Internet research using Internet devices.
    I think I will close here. Thank you for the invitation to 
testify, and I would be happy to address any further questions.
    Senator Carper. Great. Thank you. Thank you, Dr. Brown. Dr. 
Hill, please.

 TESTIMONY OF ROBERT B. HILL, PH.D.,\1\ SOCIOLOGIST AND FORMER 
    CHAIR OF THE ADVISORY COMMITTEE ON THE AFRICAN AMERICAN 
                 POPULATION, U.S. CENSUS BUREAU

    Mr. Hill. Good afternoon, Mr. Chairman and Members of this 
Subcommittee. I am pleased to be invited to provide testimony 
on this very important subject.
---------------------------------------------------------------------------
    \1\ The prepared statement of Mr. Hill appears in the Appendix on 
page 95.
---------------------------------------------------------------------------
    My testimony will focus on a major lesson learned from 
prior Censuses: The importance of developing strong 
partnerships and community outreach strategies with hard-to-
count populations in order to reduce the minority undercount in 
the Census.
    My initial experience with Decennial Censuses goes back to 
1969, when I was appointed National Director of the National 
Urban League's 1970 Census Project. This project was launched 
by Whitney Young, who was the Executive Director of the 
National Urban League at that time, and it was designed to 
reduce the black undercount in the Census. This was the first 
national partnership between the U.S. Census Bureau and a 
minority organization with over 100 branches throughout the 
country. The primary purpose of the 1970 Census Project, whose 
slogan was, ``Make Black Count,'' was to educate African 
Americans about the importance of the Census and to encourage 
them to cooperate.
    Our community outreach project was successful in convincing 
large segments of the African American community to participate 
in the 1970 Census. However, post-Census studies revealed that 
there was still a sizeable undercount of African Americans and 
other minorities in the 1970 Census. We believe that a major 
reason for the historic undercount of minority groups was the 
failure of the Census Bureau to adequately involve minority 
representatives in the advance planning and implementation of 
Decennial Censuses.
    Therefore, in his testimony to the House Census Oversight 
Committee in September 1970, Whitney Young recommended that the 
Census Bureau establish ongoing minority advisory committees to 
improve its strategies for reducing the undercount. Indeed, in 
1975 under the visionary leadership of Vincent Barabba as the 
Census Bureau Director, the first Race and Ethnic Advisory 
Committees, which are also called REACs, were formed to assist 
the Bureau in planning for the 1980 Census. The initial REACs 
comprised four minority groups: African Americans, American 
Indians, Hispanics, and Asians. For the 2000 Census, a fifth 
group was added, Native Hawaiians and Pacific Islanders.
    I served as Chair of the African American Committee in the 
planning for the 1980 Census, was reappointed for the 2000 
Census, and was involved in the initial planning for the 2010 
Census. Over the years, the Bureau has improved its methods for 
enumerating the African American population and for reducing 
the undercount among minorities. While there is still a 
differential undercount, its size has steadily declined.
    For example, while the Bureau estimated that it missed 
about 1.5 million, or 8 percent, of the black population in 
1970, it failed to count about 1 million, or 1.8 percent of 
them, in the 2000 Census. But the group with the highest 
undercount rates in Decennial Census, regardless of race or 
ethnicity, are children under 18.
    One of the most effective strategies the Bureau has used to 
reduce the minority undercount is to develop strong 
partnerships with minority groups in all phases of Census 
planning and to conduct aggressive education and outreach 
campaigns in hard-to-count communities. Based on my experience 
with prior Censuses, I would like to offer some 
recommendations.
    First, I think it is very important that Congress provides 
the Census Bureau with adequate resources to undertake the 
mammoth task of achieving a fair and accurate count. President 
Obama and Members of Congress should be congratulated for 
including an additional $1 billion in the President's stimulus 
bill to enhance the Bureau's enumeration activities in 2010. I 
was especially pleased that the bill stipulates that the Bureau 
can spend up to $250 million for its partnership program and 
outreach efforts to minority communities and hard-to-reach 
populations.
    Second, because of its comprehensive scope, the 2010 Census 
will directly stimulate this economy by hiring over half a 
million Census takers across the Nation. It is essential that 
there is an ethnically and racially diverse workforce, from the 
staff in the district offices to the enumerators in the 
neighborhoods and barrios. Members of hard-to-count populations 
should be adequately represented at all levels of Census hires, 
especially among the new partnership specialists.
    Third, one of the remarkable successes of the 2000 Census 
was the use of paid advertising to communicate messages about 
the importance of the Census to all groups. The fact that 
minority-owned advertising firms were extensively used to reach 
their respective groups played a large part in reducing the 
undercount in minority communities in the 2000 Census.
    Fourth and finally, the Bureau should permit members of its 
five Race and Ethnic Advisory Committees to play a more 
prominent role in implementing the 2010 Census, such as 
recommending partnership specialists and minority advertising 
firms, distributing foreign language Census forms, and 
identifying local sites for training Census workers and for 
serving as assistance centers to aid the elderly and other 
individuals to fill out their forms.
    These are a few suggestions I have to offer to ensure that 
the 2010 Census will be one of the most accurate and equitable 
enumerations in our history. Thank you for this opportunity.
    Senator Carper. Dr. Hill, those were great recommendations.
    I was asked in a media interview earlier today why we are 
having this hearing. One of the things, we have a 
responsibility to do is oversight. Dr. Coburn has tried very 
hard, both as the Chairman of the Subcommittee and as Ranking 
Member of this Subcommittee joined by me, to ensure we meet our 
responsibility for oversight. I replied to the reporter who 
asked me the question, ``What do you hope to accomplish from 
your hearing today,'' and what I hope to accomplish is, one, I 
would like to ask each of you to send me two names by close of 
business tomorrow of somebody that you think would be an 
excellent Director of the Bureau of the Census. By close of 
business tomorrow, give us two names of people you think are 
well equipped to do this job. I hope the Administration has 
somebody that they are vetting, that they are close to 
submitting, but just in case they don't, I want to make sure 
that we can give them a bigger talent pool to draw from.
    At least one of you said in your comments earlier today, 
talking about sort of a silver lining, high unemployment times 
right now. Well, the silver lining in that is that there is a 
great pool of talent from which to draw to work in the Census, 
whether enumerators or others, that will help reduce, I think, 
the mistakes that are made as we count people. So that is a 
potential for something good happening.
    But give us a couple of good names, each one of you, if you 
would, by close of business tomorrow. Thank you very much.
    The other thing I mentioned in response to the reporter's 
question today, I said I want to make sure that when that new 
Director of the Bureau of the Census is identified, vetted, 
nominated, confirmed, goes to work, that he or she have a 
pretty good to-do list that we have provided to him or her from 
some people who have been there and done this, not just once, 
in some cases twice and three times. We want to make sure that 
this Subcommittee is better equipped to do our job for 
oversight.
    Dr. Hill was good enough to give us four recommendations. 
Would you just run through those again real quickly, and then 
what I am going to do is ask the panel to react to those 
recommendations very briefly and to say if there are some that 
you would like to add to that. I think you have in your 
individual testimony. But beyond making sure we get an 
excellent Director on board soon, in place, just give us those 
four recommendations again, Dr. Hill.
    Mr. Hill. Essentially, the first was that the Bureau has 
adequate financial resources to conduct the Census.
    Senator Carper. And I think you said we seem to have done 
that.
    Mr. Hill. That is right.
    Senator Carper. Good.
    Mr. Hill. Second is that they should hire a workforce that 
is ethnically and racially diverse and represents the hard-to-
count and minority community populations.
    The third was that we should use the paid advertising as 
was done in 2000, but also use minority-owned firms that can 
effectively target their messages to various hard-to-reach 
groups.
    And the fourth was to permit the members of the REAC 
Committees to play a more prominent role in implementing the 
2010 Census.
    Senator Carper. OK. Talk about that last one again just a 
little bit more, please.
    Mr. Hill. These Race and Ethnic Advisory Committees are 
really very important, because they have members who come from 
diverse minority communities and work very well together. For 
example, we supported the recommendations of American Indians, 
and Hispanics and vice versa. They are also strong advocates 
for their local communities.
    One of the most effective ways of reducing the Census 
undercount is to have messages that are communicated by people 
who are trusted at the local level. That is the main 
ingredient--that local people are used who come in contact with 
others at their level, not at a higher level, and who can 
communicate the message to them. We have found this strategy to 
work every single time. These representatives can be very 
effective in many ways, such as distributing foreign language 
Census forms and helping people to complete their forms at 
assistance centers.
    All of these activities are important for reducing the non-
response follow-up. High response rates to mailed 
questionnaires will reduce the extent to which non-response 
follow-ups--which are the most tedious part of Decennial 
Censuses--are needed.
    Senator Carper. I think one of you, I don't know if it was 
Mr. Thompson, but one of you testified that the non-response, I 
think we had more people responding, was it in 2000 than in 
1990, by 2 percent? It actually went up, the number of 
respondents went up, I think you said it was 65 to 67 percent. 
Good.
    Others on the panel, if anybody would like to say that you 
think Dr. Hill has some good ideas, if you do, that is fine. 
Say that. If you think there are some other ideas on it, some 
of you made recommendations that were similar, others 
different, but I would love to hear your ideas. Let us start 
with you, Ms. Bryant.
    Ms. Bryant. I agree completely on the paid advertising.
    Senator Carper. I think you mentioned that.
    Ms. Bryant. We did not have it in 1990. The Bureau went one 
Census too long, and I am sure Mr. Thompson will agree with me, 
on depending on Public Service Announcements. The TV and radio 
stations were no longer required for their FCC licenses to give 
you around-the-clock, so we had some wonderful advertisements, 
including using four minority advertising firms, but they were 
playing at 3 o'clock in the morning. And so going to the paid 
advertising, which was a recommendation after our Census, I 
think made an enormous difference. And, of course, I am very 
attuned to Census things, but I just heard them everywhere. It 
was a really big campaign and I know it cost a lot of money, 
but you have got to pay for that.
    Senator Carper. Some of you remember the battle between the 
States and the tobacco industry a decade or so ago when the 
States attorneys general sought to extract a fair amount of 
money from the tobacco industry and a lot of that actually went 
into a foundation called the American Legacy Foundation, whose 
job it was to try to transmit to young people in this country 
the message not to smoke, not to get started, and if you are, 
stop.
    I was the founding vice chairman of that as governor at the 
time and we decided we would do a paid ad campaign. We decided 
it wouldn't be ads that guys like me would develop, but we 
would find really younger, hip ad agencies who could connect 
with young people, and they ended up putting their messages on 
TV shows that I never watched, but my sons later did. A lot of 
young people did, music stations, radio stations and so forth, 
the Internet, in ways they are just a lot smarter to connect 
that. It sounds like that is what we need to do here today.
    Ms. Bryant. Well, similarly with the African American other 
group agencies, they know the media their people watch. They 
did a great job in our Census of producing ads that just--they 
weren't given enough air time because we didn't pay for it.
    Senator Carper. Good. All right. Mr. Thompson.
    Mr. Thompson. Well, I agree with Dr. Hill. I would, as a 
footnote, note that the reason that we had such an active 
partnership program in addition to the paid advertising in 2000 
was the education that the Racial and Ethnic Advisory 
Committees provided to us on the importance of reaching out at 
the grassroots level to instill participation.
    I would, however, add a little bit to his recommendations. 
I think the most important recommendation I made, and I will 
restate it, is that there needs to be a thorough, extensive 
field test of all the systems involved in their non-response 
follow-up operation. That operation is the key to a good 
Census.
    Senator Carper. OK. Good. I think a couple of other 
witnesses have said the same thing.
    Maybe one more, Mr. Goldenkoff, and then I will yield to 
Dr. Coburn.
    Mr. Goldenkoff. Sure. I think that everything that has been 
said thus far is consistent with what GAO has said in the past, 
but I would like to put a couple of other things out on the 
table.
    Senator Carper. Please.
    Mr. Goldenkoff. Leadership. In addition to the timely 
appointment of a Census Director, what should that Census 
Director do when he or she gets in office? One, I think it is 
fairly important that the Census Director regularly reach out 
to key House and Senate committees. Transparency is key. It 
helps ensure that the Census is on track. It builds that 
confidence that we talked about earlier. It is so important, 
especially when you have an operation that is going to cost 
between $14 and $15 billion. The Director also needs to embrace 
oversight. Transparency is very critical because that also 
instills a comfort level.
    Operationally, risk and cost management are essential. We 
need to ensure that the IT systems are fully tested. That, we 
have spoken at length about. The Bureau also needs to set 
priorities.
    And I would also like to put a slightly different take on 
some of the outreach and promotion activities as well. We agree 
with the partnership specialists and need for partnerships in 
general. A Census is inherently local. But I also think the 
Bureau should consider other ways of using partners for other 
types of operations. For example, in the 2000 Census, the 
partner--I forget which locality it was, but the locality 
actually enlisted the help of their trash collectors because 
they cover the streets every single day and the locality 
trained them to look for hidden housing units. So that was a 
case where they used another local group, part of the 
government was enlisted to help in an address-building 
operation.
    Building a blog. Maybe the Census Director could consider 
putting a blog up there to deal with snafus that might come up 
or provide regular progress on the Census on a daily basis. But 
again, part of just a different take on the outreach.
    Senator Carper. That is a great list. Let me yield to Dr. 
Coburn. Thank you all for your responses.
    Senator Coburn. Well, thank you for your testimony. Mr. 
Thompson, I am not wanting to put you on the spot so I am going 
to ask this question where you don't have to answer it directly 
and then I will talk to you later. [Laughter.]
    One of the eight things you said you all did in 2000, 
number seven was that you had a strong management team in 
place. Are you in a position now where you could assess whether 
or not there is a strong management team? I am not asking 
whether there is or not. I am just saying, are you in the 
position now where you could make that assessment of what you 
see at the Census Bureau now, since you are working with them 
so closely?
    Mr. Thompson. I believe that I could provide an assessment.
    Senator Coburn. OK. I will let you off with that. I don't 
want to put you on the spot.
    Mr. Goldenkoff, you talked about risk management systems 
that need to be in place, and I have not seen your report yet, 
so I am operating at a deficit if it is out there. My staff has 
seen it and I didn't get a chance to thoroughly prep for this 
hearing. Have you all specifically listed those areas where 
they do not have now and need to have----
    Mr. Goldenkoff. I will turn it over to Mr. Powner. We did 
make 10 recommendations in the report that we issued today.
    Senator Coburn. Yes, but it is just 10 and it is in the 
report.
    Mr. Powner. Yes. It is very detailed, Dr. Coburn. There are 
six systems that need to be tested----
    Senator Coburn. I have got that down----
    Mr. Powner [continuing]. There are 44 operations----
    Senator Coburn [continuing]. And the integration----
    Mr. Powner [continuing]. And there are about 250 interfaces 
that need to be prioritized. At one time, they said they were 
going to designate a test director. We have someone who is 
working in that position part-time. We address that in the 
report. We have a recommendation for a dedicated test director. 
And then clearly we need metrics in place to really monitor 
this going forward because there is a lot to get done here.
    Senator Coburn. OK. And in your report, you are 
recommending that they have to list the areas of interfaces and 
then test them?
    Mr. Powner. Yes, absolutely.
    Senator Coburn. You make that absolute recommendation. They 
see it, and they know it is there.
    Mr. Powner. Absolutely.
    Senator Coburn. So in your opinion, if they follow your 
recommendations, both in terms of management, technical 
achievements, risk intervention, preparation for risk failure, 
testing, should they be able to accomplish what they need to 
accomplish for the 2010 Census?
    Mr. Powner. From a systems point of view, I think it is 
still highly likely they won't be able to test everything 
completely.
    Senator Coburn. Before they go into the field?
    Mr. Powner. Correct.
    Senator Coburn. How about while they are going into the 
field? In other words----
    Mr. Powner. You could continue, sure. You can continue 
while you are live, and frankly, that is what happens. I mean, 
when something goes wrong, you have bugs and you fix them on 
the fly, right?
    Senator Coburn. Right.
    Mr. Powner. But clearly, that is why the need here is to 
really prioritize. I mean, there is systems, the integration 
and the operations. One of the most important operations we 
heard, NRFU with the operational control system, that all needs 
to be tested collectively. That is clearly one of the key 
operations. But there are also others. So prioritization and 
really having these plans in place.
    But I think it is likely they won't get to some of them. 
That is why we want to see that prioritization.
    Senator Coburn. I had a conversation with a CEO of a firm, 
not this particular firm, that makes one of these. He said, in 
3 months, they could have put together a package that you could 
use for NRFU to do everything they want and transmit. They 
never were asked. Never were asked, not once. They didn't ever 
go to anything outside the contract they had. Even once they 
got in trouble, they never went to look, is there a way where 
we can still solve this, save money, have collection of data, 
transmit it. It was never asked.
    Even if we get a new Census Director, we have 20th Century 
thinking, in my estimation, at the Census Bureau, not 21st. The 
resistance to online, as Dr. Brown talked about, I mean, I have 
been fighting this for 4 years, ever since they abandoned the 
Lockheed contract. Sorry, we are just not going to do it.
    And so I have great worries. The number one worry, this 
isn't going to cost $14.8 billion. It is going to cost $18 to 
$19 billion. You wait and see. There will be another $3 or $4 
billion in an omnibus bill, emergency bill for the Census 
because, oh my gosh, we can't get it done. And part of the 
bureaucracy is don't ask for everything you need now because if 
you ask it under emergency for a Census, you are backing up 
against the window and you are going to get it. So it is going 
to cost--is it $9.3 billion, is what the 2000 Census cost in 
today's dollars, and we are going to be at least double that.
    Mr. Hill, tell me how we do--I understand the organization 
of the African American community. It is very well organized in 
a lot of these. How do I do it for Native Americans in 
Oklahoma? Tribal, yes, where we have reservations, it is much 
easier because we have an isolated group. But in States like 
Oklahoma and Tennessee and some of these other States that have 
large tribal populations but they are not reservation-based, do 
you have any ideas on how that outreach can be best 
accomplished?
    Mr. Hill. You are right about the greater difficulty of 
reaching non-reservation American Indians. There needs to be 
more aggressive outreach and targeting of community groups who 
work closely with Native Americans who do not live on 
reservations. The current approach appears to be fragmented and 
not reaching the grassroots groups. There is an urgent need to 
more effectively target community groups and tribes who work 
closely with American Indians not living on reservations. This 
strategy can work if strong local partnerships are developed 
with such indigenous groups.
    When I talk about a more prominent role for members of the 
Race and Ethnic Advisory Committees, there are many things that 
can be done by working from the bottom up. This community-based 
approach is especially needed for minorities who speak foreign 
languages.
    Senator Coburn. In other words, it is important to ask the 
question, will you come help us?
    Mr. Hill. That is right.
    Senator Coburn. Now tell us how.
    Mr. Hill. As I said before, strong partnerships can 
markedly improve outreach to the various minority groups, even 
among American Indians who do not live on reservations.
    Senator Coburn. OK. I am almost out of time. Mr. Brown, 
there was imputation used in the last Census. Would you comment 
on that, whether or not it resulted in overcount or undercount 
for where it was utilized, in your estimation, and whether or 
not that is an appropriate thing to do in this Census.
    Mr. Brown. I think imputation is essential in any Census. 
There are many situations in the field where the enumerators 
cannot talk directly to the resident of the household, and yet 
they know it is occupied. So one way or another, the data has 
to be included, from neighbors--that is part of my emphasis on 
administrative records, is that those provide potentially a 
much better method for filling in data in households you know 
are occupied.
    Senator Coburn. Do you know whether or not that would 
comply with the Supreme Court ruling?
    Mr. Brown. No, and I am not a legal expert----
    Senator Coburn. OK. Does anybody on the panel know whether 
that would comply with the Supreme Court ruling?
    Mr. Goldenkoff. I believe it does. I don't think there was 
a court decision that disputed it.
    Mr. Thompson. I believe there was--Utah raised a lawsuit 
about the use of imputation. It did go to the Supreme Court, 
and I believe that it was found to be acceptable.
    Senator Coburn. OK. Thank you. Please continue. I am sorry 
to interrupt you.
    Mr. Brown. So I do think imputation is essential. It needs 
to be done carefully. As a statistician, I think that 
statisticians could suggest a lot of improvements in the 
current imputation methodology. Much of the duplication that 
Mr. Carper mentioned is related to imputation processes, and so 
there is plenty of room to improve on this product, but I think 
it is needed in some form or other.
    Senator Coburn. Thank you.
    Mr. Chairman, you have been very generous with time. Thank 
you for holding this hearing. I would make one comment before I 
have to leave. I think it is real important that your 
Subcommittee, even if we don't----
    Senator Carper. I would say, our Subcommittee. We used to 
call it our Subcommittee.
    Senator Coburn [continuing]. Our Subcommittee, even if we 
don't have a Census Director named, whoever is acting and 
whoever is along before your committee on the basis of the 
recommendations of the GAO and also what we heard here today 
and see where the planning is ongoing.
    Senator Carper. Good. I think that is a very good 
recommendation.
    All right. Senator Burris.
    Senator Burris. Thank you, Mr. Chairman.
    Senator Coburn, your last statement was where is the Acting 
Director, would be my question, and we should certainly have 
that Acting Director here, Mr. Chairman.
    Second, Mr. Chairman, I would also give you a name of a 
person who can move in that direction in terms of running this 
operation because right now, evidently from what I hear from 
GAO, they are really behind the eight-ball in terms of what 
direction will come in order to get this off the ground in a 
timely fashion.
    But I am concerned with Dr. Hill's testimony in terms of 
participation of minority groups, and since the Acting Director 
and you all are not really from the Census Bureau, it wouldn't 
do any good to say that what Dr. Hill is saying is what is 
needed to make sure that we don't get an undercount. We need 
them involved, every group involved in every level of the 
Census steps, from the planning and the implementation, so that 
we can make sure that the Indians are counted, the Asians are 
counted, the Hispanics are counted, even the Appalachians are 
counted. We need to have everyone counted and my assessment of 
that is that there is a way of doing it and we ought to make 
sure that we also use minority contractors in order to reach 
those individuals. When all those monies are going to be spent, 
some of those dollars are going to have to be given to the 
advertisers and marketing people and their various groups.
    So any comments on that? Ms. Bryant, you ran that.
    Ms. Bryant. Well, I think our Census was the first one--
Vince Barabba is not here to argue with me--that did use 
minority advertising contractors to produce the spots. As I 
say, we did not have paid advertising, but they produced very 
excellent spots that were----
    Senator Burris. They just didn't have time to run during 
the regular----
    Ms. Bryant. Well, they ran, but they ran at 3 o'clock in 
the morning.
    Senator Burris. Yes, I heard you say that.
    Pardon me. Dr. Hill, do you want to comment on that?
    Mr. Hill. Yes. In the 2000 Census, we used paid advertising 
for the first time and it was very effective because the REAC 
Committees recommended a number of these firms. Each of the 
minority firms targeted specially-prepared messages to their 
respective groups. For example, the African American ad firm 
not only targeted messages to African Americans, but also to 
black immigrants from the Caribbean and Continental Africa, 
such as Haitians, Nigerians, Jamaicans, Trinidadians, etc. 
Thus, numerous focus groups were held with black non-immigrants 
and immigrants to develop culturally-sensitive messages.
    Senator Burris. Sure.
    Mr. Hill. The 2000 Census demonstrated that paid 
advertising works, since minority subcontractors targeted 
specially-developed messages for their groups. Paid advertising 
can be even more effective in the 2010 Census.
    Senator Burris. Are you still involved with the Census----
    Mr. Hill. No, I am not currently a member of the 
AfricanAmerican Advisory Committee. I cycled off a few years 
ago.
    Senator Burris. Who is replacing you in terms--is there a 
Dr. Hill in the Census Bureau? [Laughter.]
    Mr. Hill. I am not familiar with the current members of the 
African American Advisory Committee. But those members usually 
span a cross-section of persons who represent many sectors, 
such as ministers, community-based groups, grassroots leaders, 
etc.
    Senator Burris. OK. Then, Mr. Powers, while technology has 
led to ambitious steps to administer the Census, are there any 
new ways we can employ to combat undercounting typically 
underrepresented--any technology. Is there any technological 
emphasis that we could use?
    Mr. Powner. Well, I think clearly if you could have used 
the hand-helds for the non-response follow-up, I mean, any use 
of technology for going after the folks via non-response 
follow-up, there is potential there. We are back to paper-based 
operations there. So that is one area that you could 
potentially use----
    Senator Burris. Well, Senator Coburn just raised his cell 
phone and said that technology could be used. Is there some 
kind of way we can get that information to the Acting Director 
so that they can start looking at what technology would be 
there available for----
    Mr. Powner. What happened there, there was a plan to use 
that and then that contract ran into many problems that 
Chairman Carper looked at in great detail.
    Senator Burris. Oh, OK.
    Mr. Powner. And what we did is reverted back to things they 
knew how to do with paper. But going forward, I think that is a 
very valid point, is we need to look at using technology----
    Senator Burris. So can we look at that now? We have about 
18 months, you said, isn't it? No, 12 months----
    Mr. Powner. I think now from a risk mitigation point of 
view, you probably want to stay the course, but you want to 
look at that for 2020. We mentioned the Internet. We mentioned 
using hand-helds for more operations. I mean, we need to start 
thinking ahead for that. But right now, we are kind of in 
emergency management mode and I am not certain we would want to 
introduce that.
    Senator Burris. I mentioned the fact, too, that people 
might be afraid of their identity being stolen. Is there any 
technology that would try to ensure that when they give this 
information, it would say that it won't be given to the Social 
Security Administration or the IRS? Individuals are going to be 
a little skeptical now that their identity could be swiped so 
easily. So is there any technology that we are looking at to 
protect that?
    Mr. Powner. Well, clearly, I think there are human 
processes you want to have in place to protect identity theft. 
But from an information security point of view, all these 
systems need to be accredited and certified as secure to help 
avoid that situation.
    Senator Burris. Thank you very much. Thank you, Mr. 
Chairman. I appreciate it.
    Senator Carper. Thank you, Senator Burris.
    At least one of our witnesses mentioned the importance of 
having a set term, maybe 5 years, for a Census Director, and I 
could think of arguments for doing that and I could think of 
arguments not to do that, but I would appreciate each of our 
witnesses saying whether or not you think that is a good idea. 
I think our Commissioner for the IRS, as I recall, now serves a 
5-year term. We have some other positions where people serve 
terms, not uncommonly a 5-year term. The idea is to overlap 
from one Presidential term into the other to carry over. What 
do you all like about that idea, or on balance, what do you 
think you don't like?
    Ms. Bryant. Well, since I am the one that brought it up----
    Senator Carper. I thought you did.
    Ms. Bryant [continuing]. The problem is the 10-year cycle 
of the Decennial Census and the 5-year cycle of the Economic 
Censuses is just out of tune with a 4-year cycle for the 
current appointment. As I say, I am the worst example since I 
got into office 3 weeks before the Census.
    Senator Carper. I hope you were a quick study. I bet you 
were. You needed to be.
    Ms. Bryant. I sure worked hard at it, with the help of John 
Thompson, I might say. He was my tutor on some of the 
statistical parts.
    Senator Carper. All right. Thank you.
    Mr. Thompson, having tutored Ms. Bryant at an earlier stage 
in your life, what do you think of her idea?
    Mr. Thompson. She is being very kind about who tutored who. 
I think that it is a very good idea to have a term appointment 
for the Director of the Census Bureau. It is very unsettling 
when you have one Director leave and another Director come in. 
There is a period where you don't have leadership, and 
understanding exactly when that will happen and having a term, 
I think is very good.
    Senator Carper. Any downside?
    Mr. Thompson. I don't see any downside.
    Senator Carper. All right.
    Ms. Bryant. I will interject that this is being supported 
now by all seven past living Directors.
    Senator Carper. All right.
    Ms. Bryant. So bitter experience has told us this would be 
a good idea.
    Senator Carper. Thank you.
    Mr. Goldenkoff, do you have any thoughts?
    Mr. Goldenkoff. It has the potential to provide that 
continuity that is so important. As we have said, the Census 
cycle, it is at least a 10-year cycle, so you need someone who 
is not really going to be a temporary employee, someone who is 
in and out. And I am not even sure of all the past Census 
Directors how many of them served as long as 4 years. If so, it 
is relatively rare. And so you need someone, certainly someone 
in place who has that continuity and has longer-term vision.
    Senator Carper. Thank you. Any downsides you can think of?
    Mr. Goldenkoff. Well, to the extent that--I am thinking 
about the legislation that has been proposed to make the Census 
Bureau an independent agency, of which that is a provision. It 
doesn't necessarily follow that having a 5-year term of office 
would make the Census Bureau independent. I mean, there are 
other factors in play, and even those agencies that have a 
fixed term of office, they are just as susceptible to political 
influence, partisan influence, as agencies with a tenure that 
follows the President's. So it is also a function of the 
personality of the Director, so--it is not a panacea, is my 
point.
    Senator Carper. In the Department of Treasury, we have the 
Commissioner of the IRS, so you have a good point.
    Ms. Bryant. And Director of the Bureau of Labor Statistics, 
the other big statistical agency.
    Senator Carper. There you go. But I think maybe in each of 
those instances, certainly in IRS, the Department of Treasury, 
you have a Commissioner serving a 5-year term and the head of 
the Department is a political appointee who may not be there 
for 5 years.
    Mr. Powner, any thoughts on this?
    Mr. Powner. Yes. I think the continuity of the leadership 
is key and would be very supportive of that. We do a lot of 
work for you, Mr. Chairman, looking at the management of the IT 
budget, $70 billion spent across all Federal agencies. This has 
been looked at with Federal CIOs. So if you are a political 
appointee, CIO, your average tenure is less than 2 years. If 
you are career, it is slightly over 2 years, but still less 
than 3 years. And I think it has been well accepted in that 
that 4 to 5 years would be much better to have the continuity 
from a technology point of view and I think it applies here, 
also.
    Senator Carper. Good. Thank you. Dr. Brown.
    Mr. Brown. So let me remark on that from my perspective. 
Continuity of leadership is really very important and I think 
you mentioned, or maybe Mr. Coburn, the fact that in many 
respects, the Census Bureau is a 20th Century vehicle operating 
in the 21st Century. I think part of the problem has to do with 
the lack of continuity in leadership, both at the top of the 
Bureau and a little bit further down in terms of research and 
development.
    Senator Carper. Thank you. Dr. Hill, any thoughts on this 
idea of a 5-year term for the Director?
    Mr. Hill. I support it. Most critical decisions for the 
Census occur between 3 to 5 years before that Census. I think a 
5-year term is in a good direction, and anything that would 
make it as nonpartisan as possible is preferred.
    Senator Carper. OK. How about the idea that the Director of 
the Census should report directly to the President? I think 
that is in some legislation that the House is considering. And 
we have had it in--this Subcommittee is part of the Committee 
on Homeland Security and Governmental Affairs. Part of our 
jurisdiction is FEMA, and there has been a lot of discussion 
since Hurricane Katrina that the head of FEMA should report 
directly to, not to the Secretary of Homeland Security, but 
should report directly to the President, and that is a debate 
that is probably still going on in some circles. So it is not 
an argument that we hear just in the instance of the Census 
Bureau.
    But do we need to have, in your judgment, a situation where 
the head of the Census reports to the President as opposed to 
the Secretary of Commerce?
    Ms. Bryant. Well, having the Census Bureau as an 
independent agency, as the National Science Foundation is, for 
example, another apolitical type of organization, would remove 
two layers of bureaucracy that the Census Director or anybody 
at the Census Bureau has to go through in order to talk to you 
in Congress, to talk to the press, and I am one for flattening 
organizations.
    Senator Carper. Mr. Thompson, do you have any thoughts?
    Mr. Thompson. I really don't have an opinion on that.
    Senator Carper. Fair enough. Mr. Goldenkoff, any thoughts?
    Mr. Goldenkoff. I think there would need to be some 
safeguards in place, with the Census, it is important to have 
impartial data, and so anything, even the appearance that there 
was some type of political influence going on could really 
undermine the credibility of Census data. So I think that you 
would want some type of White House oversight or a connection 
there. It should be on the management and operations, but it 
should stop at anything that has to do with the science of 
taking the Census. So it is just finding that right balance and 
having the appropriate safeguards in place and I think that is 
really what is critical.
    Senator Carper. All right. Thank you. Mr. Powner, any 
thoughts?
    Mr. Powner. I think you could be effective with either 
scenario. I think what is most important, though, is to get the 
right leadership and the right management processes in place, 
and that has been the primary issue with the Census Bureau.
    Senator Carper. Thank you. Dr. Brown.
    Mr. Brown. I have really no thoughts about this.
    Senator Carper. OK. Dr. Hill, any thoughts?
    Mr. Hill. I agree with Mr. Goldenkoff that more balance is 
needed. I am not certain whether it should be separate or part 
of the Commerce Department. I just think that, however it is 
structured, it should have the freedom to make independent 
programmatic decisions.
    Senator Carper. All right. We have operated in the time I 
have been here in preparing for the upcoming Census where a 
Secretary of Commerce oversees his empire, which includes the 
Census Bureau, and there was a time when we realized, sitting 
in this room, the Secretary of Commerce wasn't very mindful of 
those responsibilities, his oversight responsibilities in terms 
of providing directional leadership for the Bureau of the 
Census.
    Once we got his attention and once he focused on that 
responsibility, he was a great asset, going to the 
Administration, OMB, the President, and saying, we need extra 
resources, and some of you, I think Dr. Hill and others, said 
one of the primary recommendations you have is make sure we 
have the right resources to go out and do the job well. Once 
Secretary Gutierrez, to his credit, realized we had a problem 
here, he got engaged and helped us move that and worked with 
the appropriators to make sure we had the resources that were 
believed to be necessary. So it actually can be helpful in that 
regard.
    Ms. Bryant. Well, when I talk about we are removing two 
levels of bureaucracy, that also means between the Census 
Bureau and the OMB, so another place.
    Senator Carper. All right. Thank you.
    Others have sort of talked around this question and I just 
want to come back to it one more time. I think in Senator 
Burris's questioning, he was asking Mr. Powner or Mr. 
Goldenkoff about actually acting now to better ensure that the 
technology that Dr. Coburn believes is available, that we 
actually put it to use now to get a better outcome for 2010. I 
think your response was, well, right now, let us just do what 
we said we were going to do and do that well, implement that 
well, test out the systems that still need to be tested out 
rather than starting anew.
    Any advice for us as we go through the next year or two 
doing well, counting well, accurately, and in a cost-effective 
way the people who live in this country, but is there anything 
that we ought to be mindful of doing to better ensure that when 
we get to 2020, we are not doing a Census that is at least part 
pencil and paper again?
    Mr. Goldenkoff. I would agree with some of the comments 
that were made earlier. If you look at the way the Census is 
being conducted today, with the exception of some improvements 
in technology, it is basically the same approach that has been 
used since 1970. We talked about there were some changes in 
advertising, but essentially it is a mail-out, mail-back 
operation and that approach has really exhausted its potential 
to count the Nation cost effectively. The Bureau has to spend 
more money, work harder to get essentially the same result.
    So I think that we really need to look at what other 
approaches are out there that can either get better results or 
at least control the costs better, and whether that is 
administrative records or new uses of technology, maybe some of 
the rules of the Census need to be reexamined given changes in 
society. Does it make sense to knock on a door six times during 
non-response follow-up? You have reached the point of 
diminishing returns on that. All these things probably need to 
be on the table and should be reexamined.
    Senator Carper. All right. Any other thoughts on this?
    Ms. Bryant. Well, I think a lot of people would respond on 
the Internet and we wouldn't have to do all this mail-out, 
mail-back, so forth and so on. However, I am not sure that will 
help on reducing undercount because there is a skew on who uses 
the Internet.
    Mr. Brown. That is true. There is at least one aspect that 
has been mentioned to us where it could be quite helpful 
because if you have Internet response option, it is easy to 
incorporate Hispanic, Spanish and other foreign languages----
    Ms. Bryant. Languages, yes.
    Mr. Brown [continuing]. And it also can be easier for 
proxies to help people fill out and respond. So there are ways 
in which the Internet can help, although the primary target 
population is probably--it is probably more an issue of cost 
saving than response improvement.
    Senator Carper. OK. All right. Thank you.
    Given the cost of the Census in an era of unprecedented 
fiscal challenges, what are the cost drivers of the Census and 
how can the Bureau produce an accurate yet cost-effective 
Census? Anyone at all?
    Ms. Bryant. Well, unfortunately, cost cutting was precluded 
by that Supreme Court decision because the most expensive thing 
is going after the non-respondents. The design of 2000 included 
sampling the non-respondents and estimating the rest and that 
got shot down by the Congress sending up a bill to the Supreme 
Court that the Supreme Court supported.
    Senator Carper. Well, it sounds to me like part of what you 
all have said, a number of you said, one, make sure that--the 
key here is to try to reduce the number of people that aren't 
responding. A good ad campaign properly conceived and 
implemented, I think can help on that. The idea of these 
partnerships that we talked about earlier and making sure that 
if we are interested especially and we have low response rates 
from, we will say, African Americans, from Latino Americans, 
from Native Americans, to make sure that we are involving, I 
think as Dr. Hill said, make sure that we are involving folks 
maybe who do public relations campaigns, public outreach 
campaigns, to make sure that we have included in folks who are 
formulating those campaigns folks whose background is maybe 
similar to those ethnic groups.
    Mr. Goldenkoff. I think one--I would like to add to that a 
little bit, though. We know what some of the major cost drivers 
are, non-response follow-up probably being the largest cost 
driver. One of the things that the Bureau has--can do a better 
job of is identifying where it gets the most bang for the buck. 
It has a number of repetitive operations, and it is probably a 
good thing to have some redundancy in addressing building 
operations. However, in 2000, I think there were about a dozen 
different operations to build the address list. Well, do you 
need all of them?
    I think what the Bureau has not really done a good job of 
is identifying where it gets the most results from, and that is 
true with the way it builds the address list, advertising, 
where does it get the most bang for the buck in terms of 
advertising. Is it paid advertising or is it through very 
locally-targeted partnership efforts? So maybe that is 
something the Bureau should be thinking about now, is how to 
evaluate, working on evaluation so that come 2020, they will 
have a better idea of where to invest their resources.
    Senator Carper. That is a very good point.
    Any other thoughts on this question? All right.
    I have two more, two more to go. In April, like next month, 
the Bureau is scheduled to begin its address canvassing. That 
is an operation that in total requires some 140,000 temporary 
workers who rely on hand-held computers to verify addresses and 
map information to update the Bureau's master address file and 
digital maps. And this maybe should be as much a question for 
our friends from GAO as not, but are you confident that the 
hand-helds will perform as expected?
    Mr. Goldenkoff. I think that we have more confidence than 
we did last summer. The Bureau conducted an operational field 
test back in December. It was in Fayetteville, North Carolina. 
And certainly some of the problems and issues with the data 
transmission, the unreliability, did not recur in Fayetteville. 
But what the Bureau demonstrated by that in a large degree was 
that under conditions similar to Fayetteville, the hand-helds 
will work.
    Obviously, the country is very different in respects from 
Fayetteville. There are urban areas. There are suburban areas. 
And that is the big unknown right now. And so they have made 
progress, and I think that is commendable from where they were 
back in the summer, but overall success is still an open 
question.
    Senator Carper. All right. Anybody else on this one?
    All right. The Administration's fiscal year 2010 budget 
doesn't propose increasing recruitment and hiring of Census 
takers for non-response follow-up as Congress envisioned in 
approving extra funds for the Census in the stimulus package. 
One of you mentioned, I think, $1 billion was added in the 
stimulus package for this purpose. Would the Census Bureau be 
better able to meet the challenge of lower-than-projected mail 
response if it could recruit and hire additional field staff 
heading into peak Census operations in 2010?
    Mr. Hill. Yes. We keep coming back to the non-response 
follow-up. One of the biggest obstacles to an accurate Census 
count is the high turnover of enumerators. However, the 
depressed economy should result in an influx of Census workers 
and greater stability and continuity. A major priority of the 
2010 Census should be to increase the continuity of Census 
takers and to reduce the historic high turnover rates, 
especially in inner-city areas.
    Senator Carper. Yes, sir?
    Mr. Thompson. I think this would be a really good question 
to ask the Census Bureau for the following reasons. They do 
have plans in terms of they understand how many people they 
have to recruit, how many they have to hire. They understand 
pay rates. But it would be good if they would come before you 
and explain their assumptions and rationale so that you would 
feel comfortable that they have thought through the process or 
the numbers that they have and what they have budgeted for the 
recruiting and hiring. In 2000, we had enough funding from the 
Congress that we were able to put that operation together.
    Senator Carper. All right. One last thing. Do you all have 
anything you would like to add? We had, I think, a very good 
discussion here. Anything that comes to mind that you say, oh, 
I wish I had said this? Usually, when I walk out of here, I 
think, boy, I wish I had said that or asked something else. 
Anything you all want to add as take-away? No? OK.
    A couple of thoughts. Let me just kind of wrap up what I 
think I have heard here. One is everybody says we need a first-
rate Director of the Census Bureau and we need him or her right 
now.
    Second, I think what I have heard is, for the most part, 
people say we want to make sure that whoever is leading this 
operation has the resources, human resources and financial 
resources and technology resources to do the job as best we 
can, to make sure we count as closely as we can the number of 
people who live in this country and do it in a cost-effective 
way.
    I think I have heard here that we maybe ought to give 
serious consideration to whoever is going to be serving as our 
Census Bureau Directors in the future to be nominated and 
confirmed to serve a multi-year period of time, maybe 5 years, 
maybe something more or less. Sort of a mixed bag in terms of 
whether or not we need a direct report from the Census Bureau 
directly to the White House.
    I take away from here the importance of having an ad 
campaign and a well thought-out ad campaign, particularly 
involving in the creation of the ad campaign folks who can 
better design the campaign to go after our target audience of 
people that aren't responding.
    I mentioned earlier the American Legacy Foundation where I 
was privileged to serve as their founding Vice Chairman back in 
the late 1990s, right at the turn of the century, and the folks 
who developed the advertisements to young people were not, as I 
said, adults, not for the most part. Actually, the ideas came 
from the kids. Literally, the ideas came from the teenagers and 
they worked with ad agencies, but the raw product ideas came 
from the kids. The testing was on the kids, teenagers and so 
forth, even younger than that. But that is a message or a 
lesson that I am taking away from here.
    A couple of others, as well. I won't go any further, but I 
think just a lot of good reminders here. I was talking to a 
friend of mine today about basketball and he was talking about 
more games are won in the planning of the game. I think he was 
talking about Bobby Knight. Remember Bobby Knight at the 
Indiana University, later was at Texas Tech--was it Texas Tech? 
Was that where he went? But he didn't always have the best team 
on the floor, but he always had the best plan going into a game 
of just about anybody. That is why he was so successful. I 
think having a good plan here is valuable.
    I think another good idea that we had was that a good deal 
of our operation--the program hasn't been actually tested 
operationally and we have a fair amount of work still to do 
there before we are ready for prime time. A whole bunch of good 
ideas.
    One of the things I am inclined to do, I have asked you to 
provide for me two good names by close of business tomorrow of 
folks that we could submit to the Administration as a talent 
pool in case they need some help in that regard.
    The other thing I might do, once we have got somebody who 
has been nominated, vetted, confirmed, in office, we might want 
to pull you back together again, maybe just on the phone, not 
even in person, maybe just do it on the phone in some kind of 
teleconference call to spend some time with some of us, our 
staff, the new Director, maybe a person or two from his or her 
team, just to go through again some of these points. I just 
think that might be time well spent. And if you might find time 
to do that with us, I would be grateful.
    And you are going to get a couple of questions from people 
who are Members, some who were here and some who weren't, but 
some follow-up questions. We would ask that you respond to 
those promptly. I think we are going to leave the record open 
for 2 weeks for that, so if you get any follow-up questions 
from us, please respond to them promptly.
    Are we forgetting anything here? All right. It has been a 
very good hearing, timely, and I think most informative, and we 
are grateful to you for spending your time with us and thank 
you for your ideas and input and for your willingness to give 
us some help between sundown now and sundown tomorrow, and also 
maybe your willingness to join us in a conversation with the 
new Director. Hopefully, he is going to be identified, vetted, 
confirmed soon. Thank you very much.
    With that, this hearing is adjourned.
    [Whereupon, at 4:11 p.m., the Subcommittee was adjourned.]

                            A P P E N D I X

                              ----------                              

[GRAPHIC] [TIFF OMITTED] 49639.001

[GRAPHIC] [TIFF OMITTED] 49639.002

[GRAPHIC] [TIFF OMITTED] 49639.003

[GRAPHIC] [TIFF OMITTED] 49639.004

[GRAPHIC] [TIFF OMITTED] 49639.005

[GRAPHIC] [TIFF OMITTED] 49639.006

[GRAPHIC] [TIFF OMITTED] 49639.007

[GRAPHIC] [TIFF OMITTED] 49639.008

[GRAPHIC] [TIFF OMITTED] 49639.009

[GRAPHIC] [TIFF OMITTED] 49639.010

[GRAPHIC] [TIFF OMITTED] 49639.011

[GRAPHIC] [TIFF OMITTED] 49639.012

[GRAPHIC] [TIFF OMITTED] 49639.013

[GRAPHIC] [TIFF OMITTED] 49639.014

[GRAPHIC] [TIFF OMITTED] 49639.015

[GRAPHIC] [TIFF OMITTED] 49639.016

[GRAPHIC] [TIFF OMITTED] 49639.017

[GRAPHIC] [TIFF OMITTED] 49639.018

[GRAPHIC] [TIFF OMITTED] 49639.019

[GRAPHIC] [TIFF OMITTED] 49639.020

[GRAPHIC] [TIFF OMITTED] 49639.021

[GRAPHIC] [TIFF OMITTED] 49639.022

[GRAPHIC] [TIFF OMITTED] 49639.023

[GRAPHIC] [TIFF OMITTED] 49639.024

[GRAPHIC] [TIFF OMITTED] 49639.025

[GRAPHIC] [TIFF OMITTED] 49639.026

[GRAPHIC] [TIFF OMITTED] 49639.027

[GRAPHIC] [TIFF OMITTED] 49639.028

[GRAPHIC] [TIFF OMITTED] 49639.029

[GRAPHIC] [TIFF OMITTED] 49639.030

[GRAPHIC] [TIFF OMITTED] 49639.031

[GRAPHIC] [TIFF OMITTED] 49639.032

[GRAPHIC] [TIFF OMITTED] 49639.033

[GRAPHIC] [TIFF OMITTED] 49639.034

[GRAPHIC] [TIFF OMITTED] 49639.035

[GRAPHIC] [TIFF OMITTED] 49639.036

[GRAPHIC] [TIFF OMITTED] 49639.037

[GRAPHIC] [TIFF OMITTED] 49639.038

[GRAPHIC] [TIFF OMITTED] 49639.039

[GRAPHIC] [TIFF OMITTED] 49639.040

[GRAPHIC] [TIFF OMITTED] 49639.041

[GRAPHIC] [TIFF OMITTED] 49639.042

[GRAPHIC] [TIFF OMITTED] 49639.043

[GRAPHIC] [TIFF OMITTED] 49639.044

[GRAPHIC] [TIFF OMITTED] 49639.045

[GRAPHIC] [TIFF OMITTED] 49639.046

[GRAPHIC] [TIFF OMITTED] 49639.047

[GRAPHIC] [TIFF OMITTED] 49639.048

[GRAPHIC] [TIFF OMITTED] 49639.049

[GRAPHIC] [TIFF OMITTED] 49639.050

[GRAPHIC] [TIFF OMITTED] 49639.051

[GRAPHIC] [TIFF OMITTED] 49639.052

[GRAPHIC] [TIFF OMITTED] 49639.053

[GRAPHIC] [TIFF OMITTED] 49639.054

[GRAPHIC] [TIFF OMITTED] 49639.055

[GRAPHIC] [TIFF OMITTED] 49639.056

[GRAPHIC] [TIFF OMITTED] 49639.057

[GRAPHIC] [TIFF OMITTED] 49639.058

[GRAPHIC] [TIFF OMITTED] 49639.059

[GRAPHIC] [TIFF OMITTED] 49639.060

[GRAPHIC] [TIFF OMITTED] 49639.061

[GRAPHIC] [TIFF OMITTED] 49639.062

[GRAPHIC] [TIFF OMITTED] 49639.063

[GRAPHIC] [TIFF OMITTED] 49639.064

[GRAPHIC] [TIFF OMITTED] 49639.065

                                 
