[House Hearing, 111 Congress]
[From the U.S. Government Publishing Office]


 
       THE 2010 CENSUS MASTER ADDRESS FILES: ISSUES AND CONCERNS 

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

                                HEARING

                               before the

                  SUBCOMMITTEE ON INFORMATION POLICY,
                     CENSUS, AND NATIONAL ARCHIVES

                                 of the

                         COMMITTEE ON OVERSIGHT
                         AND GOVERNMENT REFORM

                        HOUSE OF REPRESENTATIVES

                     ONE HUNDRED ELEVENTH CONGRESS

                             FIRST SESSION

                               __________

                            OCTOBER 21, 2009

                               __________

                           Serial No. 111-29

                               __________

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


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

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

54-384 PDF                       WASHINGTON : 2010 

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























              COMMITTEE ON OVERSIGHT AND GOVERNMENT REFORM

                   EDOLPHUS TOWNS, New York, Chairman
PAUL E. KANJORSKI, Pennsylvania      DARRELL E. ISSA, California
CAROLYN B. MALONEY, New York         DAN BURTON, Indiana
ELIJAH E. CUMMINGS, Maryland         JOHN L. MICA, Florida
DENNIS J. KUCINICH, Ohio             MARK E. SOUDER, Indiana
JOHN F. TIERNEY, Massachusetts       JOHN J. DUNCAN, Jr., Tennessee
WM. LACY CLAY, Missouri              MICHAEL R. TURNER, Ohio
DIANE E. WATSON, California          LYNN A. WESTMORELAND, Georgia
STEPHEN F. LYNCH, Massachusetts      PATRICK T. McHENRY, North Carolina
JIM COOPER, Tennessee                BRIAN P. BILBRAY, California
GERRY E. CONNOLLY, Virginia          JIM JORDAN, Ohio
MIKE QUIGLEY, Illinois               JEFF FLAKE, Arizona
MARCY KAPTUR, Ohio                   JEFF FORTENBERRY, Nebraska
ELEANOR HOLMES NORTON, District of   JASON CHAFFETZ, Utah
    Columbia                         AARON SCHOCK, Illinois
PATRICK J. KENNEDY, Rhode Island     BLAINE LUETKEMEYER, Missouri
DANNY K. DAVIS, Illinois             ANH ``JOSEPH'' CAO, Louisiana
CHRIS VAN HOLLEN, Maryland
HENRY CUELLAR, Texas
PAUL W. HODES, New Hampshire
CHRISTOPHER S. MURPHY, Connecticut
PETER WELCH, Vermont
BILL FOSTER, Illinois
JACKIE SPEIER, California
STEVE DRIEHAUS, Ohio
JUDY CHU, California

                      Ron Stroman, Staff Director
                Michael McCarthy, Deputy Staff Director
                      Carla Hultberg, Chief Clerk
                  Larry Brady, Minority Staff Director

   Subcommittee on Information Policy, Census, and National Archives

                   WM. LACY CLAY, Missouri, Chairman
PAUL E. KANJORSKI, Pennsylvania      PATRICK T. McHENRY, North Carolina
CAROLYN B. MALONEY, New York         LYNN A. WESTMORELAND, Georgia
ELEANOR HOLMES NORTON, District of   JOHN L. MICA, Florida
    Columbia                         JASON CHAFFETZ, Utah
DANNY K. DAVIS, Illinois
STEVE DRIEHAUS, Ohio
DIANE E. WATSON, California
                     Darryl Piggee, Staff Director












                            C O N T E N T S

                              ----------                              
                                                                   Page
Hearing held on October 21, 2009.................................     1
Statement of:
    Groves, Robert, Director, U.S. Census Bureau; Robert 
      Goldenkoff, Director, Strategic Issues, Government 
      Accountability Office; Todd Zinser, Department of Commerce, 
      Inspector General; and Ilene Jacobs, director, litigation, 
      advocacy and training, California Rural Legal Assistance, 
      Inc........................................................     8
        Goldenkoff, Robert.......................................    27
        Groves, Robert...........................................     8
        Jacobs, Ilene............................................    69
        Zinser, Todd.............................................    52
Letters, statements, etc., submitted for the record by:
    Clay, Hon. Wm. Lacy, a Representative in Congress from the 
      State of Missouri, prepared statement of...................     4
    Goldenkoff, Robert, Director, Strategic Issues, Government 
      Accountability Office, prepared statement of...............    29
    Groves, Robert, Director, U.S. Census Bureau, prepared 
      statement of...............................................    13
    Jacobs, Ilene, director, litigation, advocacy and training, 
      California Rural Legal Assistance, Inc., prepared statement 
      of.........................................................    72
    McHenry, Hon. Patrick T., a Representative in Congress from 
      the State of North Carolina, prepared statement of.........    85
    Zinser, Todd, Department of Commerce, Inspector General, 
      prepared statement of......................................    54


       THE 2010 CENSUS MASTER ADDRESS FILES: ISSUES AND CONCERNS

                              ----------                              


                      WEDNESDAY, OCTOBER 21, 2009

                  House of Representatives,
   Subcommittee on Information Policy, Census, and 
                                 National Archives,
              Committee on Oversight and Government Reform,
                                                    Washington, DC.
    The subcommittee met, pursuant to notice, at 2:25 p.m. in 
room 2154, Rayburn House Office Building, Hon. Wm. Lacy Clay 
(chairman of the subcommittee) presiding.
    Present: Representatives Clay, Maloney, Norton, Watson, 
McHenry, Westmoreland, and Chaffetz.
    Staff present: Darryl Piggee, staff director/counsel; Jean 
Gosa, clerk; Yvette Cravins, counsel; Frank Davis and Anthony 
Clark, professional staff members; Charisma Williams, staff 
assistant; Carla Hultberg, chief clerk (full committee); Adam 
Hodge, deputy press secretary (full committee); Leneal Scott, 
information systems manager (full committee); John Cuaderes, 
minority deputy staff director; Adam Fromm, minority chief 
clerk and Member liaison; and Chapin Fay, minority counsel.
    Mr. Clay. The Subcommittee on Information Policy, Census, 
and National Archives will come to order.
    Good afternoon and welcome to today's hearing entitled, 
``The 2010 Master Address Files: Issues and Concerns.'' Today's 
hearing will examine the quality and quantity of addresses 
encompassing the master address file.
    Before we begin today, I would first like to publicly 
express my condolences to the family and friends of Bill 
Sparkman. Mr. Sparkman was a Census worker found murdered in 
Kentucky. This incident is extremely unfortunate and immensely 
troubling. Census workers are doing a great civic duty for 
their country, and it is intolerable that such violations 
occur.
    Further, I would like to commend Director Groves on his 
efforts and concern for his employees. It is evident that the 
safety and well-being of Census employees are of paramount 
concern to him. I, along with this subcommittee, await swift 
justice for those responsible for such a horrendous act.
    Also I would like to, on another note, recognize a group of 
visitors here who are part of the House Democracy Partnership. 
We have 24 members of parliament from four countries this week 
for a seminar on committee operations, with an emphasis on 
organizing and holding public hearings. The visiting members 
are observing video of a hearing and meeting with House staff 
and Members to discuss the organization and conduct of 
hearings.
    I want to welcome those members of parliament here from the 
country of Kenya, as well as Peru. Welcome. Hopefully you will 
get something out of this hearing, which I am sure you will.
    Without further ado, on our panel we will hear first from 
Dr. Robert Groves, Director of the Census Bureau. Dr. Groves 
will provide the status of the Bureau's ongoing efforts to 
compile and update the master address file, including LUCA and 
its appeal process, Special Gulf Coast initiatives, address 
canvassing, and group quarter validation.
    Welcome again, Dr. Groves.
    We will then hear from Government witnesses who will 
testify and assess the compilation of the master address file. 
These witnesses will offer recommendations they believe will 
improve the Bureau's efforts.
    Our final testimony will come from a stakeholder who will 
discuss her organization's concerns about census 2010. She will 
provide her organization's actual experiences with hard-to-
count populations. She will also offer practical solutions to 
aid in the partnership between the Bureau and community-based 
organizations.
    Without objection, the Chair and ranking minority member 
will have 5 minutes to make opening statements, followed by 
opening statements not to exceed 3 minutes by any other Member 
who seeks recognition.
    Without objection, Members and witnesses may have 5 
legislative days to submit a written statement or extraneous 
materials for the record.
    Again, the purpose of today's hearing is to examine the 
master address file. This subcommittee is committed to 
reviewing the daunting and critical task of counting the 
population. This constitutionally mandated exercise has wide 
ramifications. The results are used to apportion legislative 
districts at the Federal and State level. Moreover, the 
distribution of more than $400 billion annually in Federal 
assistance to local, State, territorial, and tribal governments 
rely upon this count. Civic prestige, marketability, and 
regional political power also rest upon these numbers.
    As we look forward to April 1, 2010, the subcommittee 
evaluates the status of the Bureau's efforts to count all 
inhabitants of this country. The master address file is an 
essential component of the 2010 decennial census; thus, an 
assessment of the compilation of our addresses is of 
fundamental interest and concern to the subcommittee.
    Today's hearing will focus on the Bureau's progress in the 
compilation, scheduling, cost, and transparency of the master 
address file. The subcommittee will explore all aspects of 
master address file, including but not limited to LUCA, the 
LUCA appeal process, address canvassing, update leave, special 
Gulf Coast initiatives, and budgetary matters. The Bureau's 
interaction and cooperation with local and county governments, 
community organizations, stakeholders will further be explored. 
The success of the census is dependent on the quality of the 
address list.
    I thank the witnesses for appearing today and look forward 
to their testimony.
    I now yield to the distinguished ranking minority member, 
Mr. Chaffetz of Utah.
    [The prepared statement of Hon. Wm. Lacy Clay follows:]

    [GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
    
    Mr. Chaffetz. Thank you, Mr. Chairman. Thank you for all of 
you who are participating in this, and thank you very much for 
our guests. I hope you find this informative and I appreciate 
your joining us here today.
    As we work on this very daunting, huge, massive task of 
trying to count every person in the United States of America 
our time line is short. Here we have less than 6 months to go, 
and undoubtedly the challenges will be huge. Of particular 
concern that I would like to make sure of is are we up to the 
task of making sure that everyone is fairly counted, that they 
are not under-counted and they are not over-counted, and that 
we have the tools necessary and the technology in place in 
order to make that happen.
    I have specific questions and concerns about the viability 
of the work force that is being hired in order to do this 
enumeration; specifically, the practice of hiring known 
criminals. I know the background checks and the fingerprinting 
have been an issue, and I would appreciate an update along the 
way in that regard.
    And then I also have questions as to why we don't or to 
what degree we do utilize and tap into what we already do in 
the U.S. Postal Service. They already have a Federal work force 
of hundreds of thousands of people who go to every home, every 
address in this country. I recognize some have Post Office 
boxes and what not, and there are certain challenges with 
illegally subdivided homes and what not, as Ms. Jacobs I am 
sure will address, but, with that being said, mapping out this 
country, why we are not more closely aligned with literally 
hundreds of thousands of people who do this on an almost 
everyday basis is something I think is worth at least a few 
questions along the way.
    I look forward to your testimony and appreciate the 
interaction today. That is what this process is all about.
    I thank the chairman for the time and yield back.
    Mr. Clay. I thank the gentleman from Utah.
    Would Mr. Westmoreland care to make an opening?
    Mr. Westmoreland. No, thank you.
    Mr. Clay. Thank you. Then, if there are no additional 
opening statements, we will now receive testimony from the 
witnesses before us today.
    I want to start by introducing our panel. We will hear 
first from Dr. Robert Groves, Director of the Census Bureau. 
Dr. Groves has authored numerous books and articles. He was a 
recipient of the prestigious Julius Shiskin Memorial Award in 
2008. He has a B.A. from Dartmouth and a master's in sociology 
and statistics, and further earned a doctorate from Michigan. 
Dr. Groves' book, ``Non-response in Household Interview 
Surveys,'' with Mick Cooper received the 2008 AAPOR Book Award.
    Dr. Groves began his tenure as Director on July 15, 2009.
    Next we will hear from Mr. Robert Goldenkoff, who currently 
serves as the Director of Strategic Issues at the Government 
Accountability Office. Mr. Goldenkoff is responsible for 
reviewing the 2010 census and Government-wide human capital 
reform. He has also developed a body of work related to 
transportation security, combating human trafficking, and 
Federal statistical programs.
    Mr. Goldenkoff's various works have been published in the 
Public Administration Review Policy Studies Journal, Government 
Executive, and Technology Review.
    Thank you for being here.
    Next we will hear from Mr. Todd Zinser, Inspector General 
of the Department of Commerce. As the Inspector General, Mr. 
Zinser leads a team of auditors, evaluators, investigators, 
attorneys, and administrative staff responsible for promoting 
economy and efficiency in detecting and preventing fraud, 
waste, and abuse in a vast array of business, scientific, 
economic, and environmental programs administered by the 
Department and its 13 bureaus.
    Thank you, Mr. Zinser, for coming today.
    Our final witness will be Ms. Ilene Jacobs from California 
Rural Legal Assistance. Ms. Jacobs is the Director of 
Litigation, Advocacy and Training. She has spent 30 years of 
her legal career as an advocate for housing and civil rights in 
low income communities in urban and rural United States. Ms. 
Jacobs taught housing law for the UC Davis Law School and women 
and the law for Yuba Community College. She obtained her B.A. 
from Boston University and J.D. from the Northwestern 
University School of Law. She has co-authored two publications 
on the under-counted farm workers and indigenous groups in the 
census.
    Ms. Jacobs is the CRLA delegate to the National 2010 Census 
Advisory Committee, for which she is chair of an ad hoc 
subcommittee on hard-to-locate housing units.
    Thank you all for appearing before this subcommittee today.
    It is the policy of the subcommittee that all witnesses 
before they testify be sworn in. Can I ask you to stand?
    [Witnesses sworn.]
    Mr. Clay. Let the record reflect that the witnesses 
answered in the affirmative.
    Each of you will have an opportunity to make an opening 
statement. Your complete written testimony will be included in 
the hearing record. The yellow light will indicate that it will 
be time to sum up, and the red light will indicate that your 
time has expired.
    Dr. Groves, you may proceed first.

  STATEMENTS OF ROBERT GROVES, DIRECTOR, U.S. CENSUS BUREAU; 
   ROBERT GOLDENKOFF, DIRECTOR, STRATEGIC ISSUES, GOVERNMENT 
  ACCOUNTABILITY OFFICE; TODD ZINSER, DEPARTMENT OF COMMERCE, 
  INSPECTOR GENERAL; AND ILENE JACOBS, DIRECTOR, LITIGATION, 
 ADVOCACY AND TRAINING, CALIFORNIA RURAL LEGAL ASSISTANCE, INC.

                   STATEMENT OF ROBERT GROVES

    Dr. Groves. Thank you. Chairman Clay, Ranking Member 
Chaffetz, other subcommittee members, I am happy to be here to 
testify with regard to the master address file for the 2010 
census.
    When I testified here on September 22nd, I noted at that 
time a set of professional judgments about the assessment of 
the 2010 preparations, and I noted at the end of this month we 
would be finished with our internal evaluation of the master 
address file. I realize the schedules of the committee didn't 
work to hit that time exactly right. I will tell you as much as 
we know, I promise today, but there is still work to be done, 
and I would be happy to meet with the full committee or any 
subset when we have that full report ready.
    Let me begin by reminding us of what the master address 
file is. It is literally an inventory of all the addresses and 
descriptions of units, along with their geographical locations. 
It is the source of the mailing of all the questionnaires and 
delivery of the decennial forms, so it is a big deal for the 
2010 census, and the quality of that master address file is 
appropriately a target of this subcommittee's scrutiny.
    There are three major quality criteria that I will talk 
about today. One is its completeness, its coverage, does it 
contain all of the housing units in the United States? Second, 
are the addresses on each of those housing units complete, or 
the physical descriptions? Can we mail or find the housing 
units in our later operations? And then third, do we know where 
these units are? Is the spatial accuracy what we need to have 
for a successful census?
    We have done three important things over the decade I think 
it is important to know that were designed to improve the 
quality of the master address file. The first thing that was 
done is a reflection of what happened in 2000. It was a result 
of the 2000 census that there were more duplicates in this 
frame than were expected. One source of the duplicates had to 
do with group quarters housing units, and we have blended those 
two lists together with the hope that will reduce the kind of 
duplication we found in 2000.
    We have realigned all the streets and roads in the country 
to reflect changes over the decade, and we are fulling using, 
in cooperation with the Postal Service, codes that determine 
how best to get forms to particular addresses, whether we 
should mail them or deliver them ourselves.
    We have also been updating this frame throughout the decade 
through, again, a cooperation with the Postal Service on the 
delivery sequence file, and then through our own field work in 
the American community survey and other surveys, especially in 
rural areas.
    Mr. Chairman, you mentioned the so-called LUCA program, 
which is the local update of census addresses. It plays, as you 
know, a critical role. It is a key, both symbolic and real, 
cooperation with local and State governments throughout the 
country, as well as tribal governments. This is an important 
part of building the master address file.
    Mr. Chairman, I want to thank you, along with 
Representative Maloney, former ranking member of the 
subcommittee, Michael Turner, Representative Michael Thompson, 
and a former ranking member of the full committee, 
Representative Tom Davis, for your support last year. You wrote 
a Dear Colleague letter in March and you helped to spread the 
word about the LUCA program to your colleagues on the Hill, and 
it helped, as you will see in just a few minutes.
    We invited in 2007 about 40,000 different tribal, State, 
and local governments to participate in this local program. 
About 12,000, 11,500 of them, registered. About 29 percent of 
the governments are represented by that 11,000. That is a 
disappointing number when you first see that, but those 
governments represent about 92 percent of the housing units in 
the country. So that is kind of the first evaluation of LUCA 
that one could mount.
    One way to evaluate it is to compare it to what happened in 
2000. In 2000 we had about 18,000 governments registered. They 
represented less than 92 percent of the total housing units in 
the country. So overall on participation in LUCA we had greater 
participation this decade than last, and that is something that 
we are grateful for.
    We received submissions reflecting changes to our address 
list from those local governments from about 79 percent of the 
governments that had registered. That is about 8,100 
governments. This compares to about 67 percent submission in 
the 1998 LUCA and 48 percent in the 1999 LUCA. So once again 
the participation, the submission of these lists to us from the 
local governments was somewhat better than in 2000, another 
good sign.
    We then matched these addresses supplied by local 
governments against the master address file and we sent out all 
those addresses for the address canvassing operation that took 
place in the summer.
    Let me mention a couple of things about other improvements 
in the local update program. We had a single cycle of review. 
This reduced the complexity of participating governments. We 
had a longer review period, 120 days versus 90. We allowed a 
variety of ways to participate that seemed to fit the different 
problems local governments were facing. We provided easy to use 
software that they could download on their desktops to help. 
And for the first time we allowed State governments to 
represent lower governments within their States. That partially 
explains the lower count of participation of governments but 
the higher percentage of housing units represented.
    The preliminary figures from this program show that about 8 
million addresses were provisionally added to the master 
address file for verification. Thirty million of the addresses 
submitted by the local update program matched addresses already 
on the file, and we had 2 million corrections to addresses.
    Then, as you know, over the summer in 2009 Census Bureau 
staff walked every street and road in the country and visited 
145 million locations that consisted of the 145 million units 
on the then master address file. The only areas that we omitted 
from this were remote Alaska and parts of Maine that represent 
about 35,000 households of the 134 million.
    I can give you the results of the address canvassing work. 
About 98 million addresses on that list of 145 million were 
verified as is. Twenty million were corrected. Usually that was 
a street name correction, small changes. Five million were 
moved to another block. Ten million were added. They weren't on 
the address list before and they weren't on the LUCA submission 
list. So at the current time, the master address file consists 
of about 134 million records. That turns out to match 
independent estimates of the housing unit count. That is a good 
sign so far in our comparisons.
    The figures show that about 21 million addresses fall into 
either a delete or duplicate or non-residential category. 
Sixteen million of those were deletes; that is, we couldn't 
find them when we got out there. And about 4 million were 
duplicates that were found to exist in other places in the 
master address file. About a million of the addresses we had on 
the list were non-residential. They have now been omitted.
    Two million of them were what we called other living 
quarters. These are things like assisted living facilities, 
dormitories, group homes. We sort of set those aside, and right 
now as we speak there are people visiting those group quarters, 
because that was a problem in 2000, making sure we had all the 
unit identifications right in those group facilities. We are 
out there right now making sure we have correct unit 
identifiers, and that is going to pay off come spring when we 
do the measurement.
    With respect to the results of address canvassing on the 
local update cases, themselves, our initial results show that 
about 66 percent of the LUCA addresses were deleted, identified 
as duplicate, or found to be non-residential. About 29 percent 
of the addresses were verified, corrected, or moved, and about 
5 percent were unresolved in address canvassing but will remain 
in the enumeration universe.
    We are now reviewing this operation, as I said at the 
opening, and I am happy to get back to you when we have all of 
the evaluative results on that program.
    You know that we have other programs that will improve 
hopefully the master address file. We are right now out asking 
the same local governments to give us new construction updates. 
We have invited about 29,000 governments to participate in 
this. About 15,000 have already said yes, and we are off and 
running on that.
    We are also going to make other updates to this file. We 
are not through for 2010. We will get other updates from the 
U.S. Postal Service, from the delivery sequence file. We are 
going to have a count review program that is going to go out 
early in 2010, and then we are going to have updates from other 
field operations. Our attempt in this is to get the most up-to-
date master address file we can.
    So let me sum up. I said there are three evaluative 
criteria for the master address file. Coverage is first. 
Relative to 2000 I noted that fewer governments participated in 
the local update program but they represented a higher 
proportion of all addresses in the country than 2000. I noted 
that State and local governments provided addresses that form 
about 2 percent of the total valid addresses on the file after 
address canvassing. And after address canvassing the total 
number of units on the file is comparable to an independent 
estimate of the count of housing units in the country.
    The second criteria is the completeness of addresses, and 
we found about 2 million other living quarters that are now 
being revisited to get those addresses right, those identifiers 
right. We are expecting a lot of these to revert to a single 
housing unit by the time we are through with this operation.
    We continue to evaluate the current status of the master 
address file, and I am hopeful that I could talk more about 
this in a later hearing in front of this committee.
    I thank the committee for this opportunity to testify and 
look forward to your questions.
    [The prepared statement of Dr. Groves follows:]

    [GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
    
    Mr. Clay. Thank you, Dr. Groves. We know you hit the ground 
running at the Census Bureau, and we thank you for your 
service.
    Mr. Goldenkoff, you are recognized for 5 minutes.

                 STATEMENT OF ROBERT GOLDENKOFF

    Mr. Goldenkoff. Mr. Chairman, Ranking Member McHenry, 
members of the subcommittee, I am pleased to be here today to 
discuss the Census Bureau's progress in building a complete and 
accurate address list. As you know, a quality address list, 
along with precise maps, are key to a successful census. If the 
Bureau's address lists and maps are inaccurate, people can be 
missed, counted more than once, or included in the wrong 
location.
    That said, compiling an accurate address list is no easy 
task. One reason for this is that people can reside in hidden 
and non-traditional housing units such as converted attics and 
basements, as well as in cars, boats, trailers, motels, tent 
cities, and labor camps.
    While these living arrangements have long existed, the 
large number of foreclosures the Nation has recently 
experienced, as well as the natural disasters that have hit the 
Gulf Coast and other regions, have likely increased the number 
of people living in less conventional housing.
    In addition to housing units, which include single-family 
homes, apartments, and mobile homes, many people also reside in 
facilities called group quarters, which include prisons, 
dormitories, nursing homes, and similar locations.
    The Bureau's database of the Nation's roughly 140 million 
addresses is called the master address file [MAF]. As 
requested, my testimony will describe the preliminary results 
of three MAF building operations that can help locate hidden 
housing units and other traditionally hard-to-count 
populations. The operations we reviewed are LUCA, address 
canvassing, and group quarters validation.
    I will also provide an update on the IT systems the Bureau 
will use to update and extract information from the MAF 
database.
    My testimony today has two main points. First, the Bureau 
goes to great lengths to ensure the accuracy of the address 
file using multiple operations that include partnerships with 
the Postal Service, extensive field verifications, and numerous 
other activities. Second, the operations we reviewed generally 
proceeded as planned and we did not observe any significant 
operational setbacks.
    Still, the overall effectiveness of the Bureau's efforts 
will not be known until later in the census when the Bureau 
completes various assessments.
    Turning first to LUCA, the Bureau partnered with State, 
local, and tribal governments, tapping into their knowledge of 
local populations and housing conditions in order to develop a 
more complete and accurate address list. More than 8,000 
jurisdictions participated in the program between November 2007 
and March 2008. However, LUCA submissions generated a 
relatively small percentage of additions to the MAF. For 
example, of around 36 million potential additions that 
localities submitted, just 2.4 million or 7 percent were new 
addresses not already in the MAF. The others were duplicate 
addresses, nonexistent or nonresidential.
    Address canvassing finished ahead of schedule, in part 
because of improvements the Bureau made to the hand-held 
computers used to collect data, as well as because of lower 
than expected employee turnover. Nevertheless, the operation 
exceeded its original budget estimate of $356 million by $88 
million, a cost overrun of 25 percent.
    A key reason for the overrun was that the Bureau did not 
update its cost estimates to reflect changes to the address 
canvassing workload. Further, the Bureau did not follow its 
staffing strategy and hired too many listers.
    Recognizing the difficulties associated with address 
canvassing in the hurricane affected areas along the Gulf 
Coast, the Bureau developed supplemental training materials to 
help listers identify addresses where people are or may be 
living when census questionnaires are distributed early next 
year. For example, the materials noted that people might be 
living in trailers, homes marked for demolition, and 
nonresidential spaces such as storage areas above restaurants.
    To help ensure group quarters are accurately included in 
the census, the Bureau is conducting an operation called group 
quarters validation, which is going on right now. The Census 
Bureau developed and tested new procedures to improve how it 
identifies and counts these facilities based on lessons learned 
from the 2000 census.
    With respect to the automated system that supports the MAF, 
although the Bureau has improved aspects of its IT management, 
we continue to be concerned about the lack of finalized test 
plans, incomplete metrics to gauge progress, and an aggressive 
testing and implementation schedule going forward.
    In summary, the Bureau has taken extraordinary measures to 
produce a quality address list and associated maps. Still, 
accurately locating each and every dwelling in the Nation is an 
inherently challenging endeavor, and the overall quality of the 
Bureau's address list will not be known until later in the 
census when the Bureau completes the assessments that Dr. 
Groves mentioned.
    Chairman Clay, Ranking Member McHenry, and members of the 
subcommittee, this concludes my remarks. I will be happy to 
answer any questions that you may have.
    [The prepared statement of Mr. Goldenkoff follows:]

    [GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
    
    Mr. Clay. Thank you so much, Mr. Goldenkoff.
    Mr. Zinser, you are recognized for 5 minutes.

                    STATEMENT OF TODD ZINSER

    Ms. Zinser. Thank you, Mr. Chairman, Ranking Member 
McHenry, members of the subcommittee. We appreciate the 
opportunity to testify today about the Census Bureau's master 
address file.
    My testimony covers three points, Mr. Chairman.
    First, building the master address file is an enormously 
important and enormously complex undertaking. Second, my office 
has focused a lot of our work on the Census operations used to 
build the file, and, not surprisingly given the enormity of the 
task, the operations are prone to errors and omissions. Third, 
Census continues to work very hard to carry out operations to 
improve the master address file and compensate for those errors 
and omissions, and the execution of those operations is 
critically important.
    The Census Bureau describes an accurate, comprehensive, and 
timely list as one of the best predictors of a successful 
census. Errors on the master address file can cause people to 
be missed or counted more than once, as well as increase cost 
and the public burden by requiring enumerators to visit non-
existent or duplicate locations during the non-response 
followup operation.
    After the 2000 census the Bureau launched an ambitious plan 
to maintain and update both the master address file and the 
census maps through a variety of operations. They accomplished 
some of their plans, but still relied on a massive address 
canvassing operation at the end of the decade as the primary 
operation for verifying, updating, or deleting addresses, 
adding missing addresses, updating streets on the maps, and 
geo-coding every structure. Address canvassing employed 140,000 
temporary workers and cost over $400 million, not including the 
cost of the hand-held computers.
    Our work over the decade on the master address file has 
identified consistent problems. We observed the 2006 site test 
in Austin and Cheyenne River Reservation, the 2008 dress 
rehearsal, and the address canvassing operation, itself. My 
written statement includes examples of the types of errors and 
omission we observed in those operations, including missing, 
duplicate, and erroneous master address file addresses, 
problems with listers not following procedures, and other 
problems with process and execution. Our work in this area has 
caused us to have concerns about the overall quality of the 
address list.
    The Census Bureau has operations designed to update the 
2009 address list and potentially mitigate some of the issues 
that affect master address file reliability. These include 
group quarters validation, the LUCA appeals process, new 
construction adds, and update leave and update enumerate. These 
represent some of the most significant efforts planned and 
underway to strengthen and update the master address file. It 
is critically important that the Bureau execute these 
operations well. I believe the Census is working hard to do so.
    However, if we were to be asked what other actions the 
Census Bureau could take at this point, we would offer two 
suggestions that could assess the quality of the master address 
file right now and provide information that could be used in 
subsequent improvement operations and potentially provide the 
opportunity for additional address list corrections.
    First, Census should take advantage of housing unit 
estimates to help assess master address file quality. For the 
past 20 years, the Bureau has produced annual estimates of 
housing units for States, counties, and local governments. 
These statistics are used as controls for several Census Bureau 
surveys and could be used for the decennial as benchmarks 
against which potential over- or under-counting of housing 
could be measured. For example, after census 2000, count 
comparisons for over 800 of the Nation's most rural counties 
indicated potential under-coverage in 275 of the counties. Use 
of housing estimates could identify these types of 
discrepancies now, before the decennial census, and perhaps 
steps could be taken to address them.
    Second, the Bureau could make greater use of administrative 
records as another source for checking address quality. Such 
records collected by all levels of Government and the private 
sector are used by census in conducting several of its 
statistical operations. By matching current administrative 
records to the master address file, Census could both assess 
master address file quality and potentially add missing 
addresses.
    Even if the Bureau determines that incorporating missed 
addresses identified in this process would not be feasible at 
this stage of the decennial, there would still be benefits to 
assessing the address file and identifying areas where 
addresses are missing. It would allow subsequent field 
operations to be alerted on a targeted basis of the high 
potential for an accurate list for that area and the need for 
greater attention to those areas.
    Mr. Chairman, this completes my summary, and I would be 
happy to respond to any questions.
    [The prepared statement of Mr. Zinser follows:]

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

                   STATEMENT OF ILENE JACOBS

    Ms. Jacobs. Thank you, Mr. Chairman, and thank you, Ranking 
Member McHenry and members of the subcommittee, for giving me 
this opportunity to speak to you today about the master address 
file and to recommend improvements with respect to the 
development of the master address file.
    I was glad to hear Dr. Groves refer to the master address 
file as a big deal, and I would like to reiterate his 
statement. It is a very big deal, particularly to the very 
hard-to-count populations that we represent throughout the 21 
rural offices of California Rural Legal Assistance. Our mission 
is to ameliorate rural poverty and to ensure that rural 
communities have access to justice and the provision of basic 
human rights; therefore, we very much understand the importance 
of an accurate census and having accurate census data, which 
starts with an accurate master address file.
    I would like to speak about the adverse impact in the 
communities we represent of having a less-than-accurate master 
address file and make some very practical recommendations for 
improvements.
    We represent the hardest-to-count populations in the 
hardest-to-locate housing. We represent farm workers, the rural 
poor, immigrants, very diverse racial and ethnic groups, 
linguistically isolated populations, elderly, disabled, and 
most recently many foreclosure victims throughout rural 
California.
    An inaccurate master address file in a mail-out/mail-back 
census means that our clients do not have adequate housing, 
they lack health care, they don't have job training, they have 
fewer educational opportunities, lower literacy, they have 
fewer needed municipal services like basic water and sewer, 
they lack community and economic development programs and 
resources, and it is harder to enforce the fundamental rights 
that they have guaranteed by law.
    The direct impact on local government is very serious, 
because they lack the ability to meet the pressing needs of the 
hardest-to-count populations. There are social and economic 
costs not only to our clients but to the local governments that 
are there to serve them.
    My prepared statement describes the structural bias in the 
development of the master address file. My involvement with the 
Census Bureau and concern about the master address file started 
when, after the 1990 census, we participated in a study that 
measured an at least 50 percent what we called mega-under-count 
of migrant and seasonal farm workers, and we attribute much of 
that mega-under-count to missing housing units. Not 
exclusively, but a significant part of that.
    The structural bias in the development of the master 
address file has not been solved. I give a lot of credit to the 
Census Bureau for making improvements in instructions to 
address listers about hidden housing units and what types of 
units to include in the address file, but we still have the 
problem of complete omission of entire households because, in 
the hardest-to-count areas and areas of high concentration of 
hidden housing units, the master address file is incomplete.
    And if the master address file for 2010 is incomplete, that 
means that the master address file for use in the American 
community survey, Census Bureau's replacement for the long 
form, also will be incomplete, and that will result not only in 
omission of housing units but omission of people, and it will 
carry into the American community survey a skewed set of 
demographic characteristics of the most needy populations, 
particularly diverse racial and ethnic groups and the hardest-
to-count populations that we represent.
    I have made five key points and five key recommendations in 
our written testimony.
    First, the hard-to-locate housing units in rural California 
and elsewhere need to be understood. They are backyard chicken 
coops, they are illegal garages, they are tool sheds that are 
rented out to families to live in, they are single family units 
and apartments that are subdivided into essentially one-room-
per-complex household, and that can be a family or an extended 
family per room in a six-apartment dwelling.
    They are motel rooms that are occupied by 20 migrant and 
seasonal farm workers at a time. They are trailer encampments. 
They are tarps and lean-tos built into canyons and the kind of 
housing that is unacceptable in this country, but, 
nevertheless, it is spread throughout rural California, and 
there are many similarities to concentrated urban areas when we 
talk about these hidden and illegal housing units.
    In our study after the 2000 census--and I refer to that in 
my written testimony--we found that a very high percentage of 
these types of units were missed in the seven communities that 
we evaluated using the Census Bureau's methodology.
    Second, address canvassing does not adequately identify 
these units, albeit improvements have been made.
    Third, ultimately, as I said, it skews the population 
profile because the hidden units tend to be occupied by the 
hardest-to-count populations who then become very difficult to 
profile in the American community survey. This has a direct 
impact on all of our communities throughout the country, and it 
can be addressed in 2010 and in the ACS.
    We first recommend that the Census Bureau adopt the address 
listing protocol that we used in 2000 in the L.A. region to 
count migrant and seasonal farm worker units and hidden units, 
that they implement this address listing protocol which was 
recognized by the GAO in its report on farm workers, and would 
extend to other hard-to-count populations.
    Second, that the Census Bureau work with the regional 
offices, census partners, community-based organizations, and 
local governments to identify areas of high concentrations of 
hard-to-locate housing units and target those for tool kit 
enumeration operations, and by that I mean specifically 
utilizing update leave and update enumerate operations within 
mail-out/mail-back areas, not only in remote areas, because 
that is where the update leave and update enumerate type 
operations are utilized. I think they can be effectively 
utilized within hard-to-count and mail-out/mail-back areas.
    We also should be using the knowledge of community-based 
organizations in the LUCA process, which presently the Census 
Bureau does not do, and the master address file should be 
evaluated in 2010 in areas of a high concentration of hard-to-
locate housing units. That evaluation then could be used to 
carry over the best practices into the American communities 
survey.
    I know I have run out of time and I apologize for that. 
There is a lot more I could say but I won't. I thank you very 
much for the opportunity to testify.
    [The prepared statement of Ms. Jacobs follows:]

    [GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]
    
    Mr. Clay. Perhaps we will have questions, but thank you, 
Ms. Jacobs, and thank the entire panel for their testimony.
    I will now go to the ranking member, who has an opening 
statement, and you can also proceed into your questions.
    Mr. McHenry. Thank you, Mr. Chairman.
    Mr. Clay. You are welcome.
    Mr. McHenry. And thank you all for coming back and 
testifying, Dr. Groves, Mr. Goldenkoff in particular, who have 
been regular in being with us, and I certainly appreciate that.
    Dr. Groves, thank you for following through on your 
commitment to testify frequently and keep us apprised of the 
process. I do appreciate all efforts that you have made. But I 
do think there has been some troubling news that we have seen 
since the Senate hearing on October 7th, and that this 
committee was not previously made aware of. First, the cost 
overruns, the significant cost overruns, I know they are not 
available, the cost estimates for address canvassing in your 
last testimony, Dr. Groves, but certainly 25 percent, going 
over budget by 25 percent is very significant.
    Second, although Dr. Groves had informed us at the last 
census hearing that better cost estimation and control was 
needed, the Bureau admitted on October 7th that its models were 
grossly inadequate and can translate into future budgetary 
problems if not immediately addressed.
    Third is the GAO study found that the Bureau's 
fingerprinting process for temporary workers was deeply flawed 
and could potentially result in criminals being hired as 
enumerators.
    While I'm disappointed that these issues were not brought 
to our attention, I have no doubt that the Bureau is actively 
working to go through and create plans, operating procedures, 
and budgets that are accurate and proper.
    But, having said that, there are some successes, as the GAO 
report indicates and as Dr. Groves' testimony indicates, as 
well. The timely and comprehensive completion of address 
canvassing was certainly a huge success, and the partnership 
programs, the media campaign efforts have been conducted in an 
unprecedented level, as no previous census has seen, reaching 
out to diverse groups of people across this country. And the 
first major wave of recruitment has met with an applicant pool 
that was much larger and more qualified than expected, although 
the GAO outlined some challenges with that, as well.
    The Bureau has also started with a 2010 local update of 
census addresses [LUCA], as you all have testified to, that has 
been the most effective to date. That is certainly good news, 
and updating the master address file with that information is 
certainly good.
    I want to reiterate my commitment to ensure that the Bureau 
stays on track with its planning, execution of the 2010 census.
    Dr. Groves, you should not limit your communication on 
issues of concern to just public hearings. We would certainly 
appreciate whatever updates you can give us so that certain 
things like the budget overrun, we don't have to find out about 
through newspapers.
    As Chairman Clay has said and I will reiterate, our doors 
are open. I think you will find that not just the ranking 
member and the chairman on this committee have their doors 
open, but all the committee members. We want to make sure that 
this is the most accurate census in our Nation's history, and I 
think we have the capacity to do that. As Mr. Mesenborg and Dr. 
Groves have testified, that is their intent and the Bureau's 
intent.
    So thank you for your testimony. I look forward to hearing 
your answers to questions.
    [The prepared statement of Hon. Patrick T. McHenry 
follows:]

[GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]

    Mr. McHenry. With that, Mr. Chairman, would you like me to 
just go right into questions?
    Mr. Clay. Yes.
    Mr. McHenry. OK. Thank you.
    Dr. Groves, rehiring the temporary workers that address the 
address canvassing, I know that is certainly what you testified 
to before. In terms of fingerprinting, what procedure will 
these rehires go through?
    Dr. Groves. I want to make sure I got the question. Are you 
looking forward to non-response followup?
    Mr. McHenry. Yes. I'm sorry. Going forward, when 
approximately 570,000, 600,000 are hired for the non-response 
followup, you will be taking applicants from the original 
address canvassing pool first, of course, because they have 
gone through the Bureau training. What is the procedure to 
check their criminal records?
    Dr. Groves. Well, let me step back a couple of steps to 
make sure that I am answering your question fully. The 
procedures on the fingerprinting are going through a critical 
review right now.
    Mr. McHenry. Yes.
    Dr. Groves. We are going to change some of those 
procedures, with the aspiration that we reduce the problem that 
we found in address canvassing, which was a failure to read 
fingerprints from some people.
    I want to remind the committee that the process by which we 
hire someone involves, first, their submission and verification 
of IDs that provide a Social Security number, a name, sex, and 
we submit that to the FBI. No one has worked on the decennial 
census nor will work on the decennial census without passing 
that name check. That will remain true.
    In 2000 only that check was used, and now we are adding 
this fingerprinting process.
    Mr. McHenry. Well, with adding the fingerprinting 
procedure--and I am sure there are other questions about this 
procedure, and we will get to that--but will all these folks 
that are re-hired for non-response followup, will they be re-
fingerprinted or will you use their original fingerprint and 
resubmit it?
    Dr. Groves. It is our current intention for those who had 
fingerprints read, submitted, accepted by the FBI, we will use 
those prints. For those who didn't have reads on their 
fingerprints, we will again go through the fingerprinting 
process.
    Mr. McHenry. OK.
    Dr. Groves. Now, I should note that not all of them will be 
re-hired. Many of them have gotten other jobs and so on.
    Mr. McHenry. Yes.
    Dr. Groves. The exact proportion of rehires versus new 
hires isn't really known at this point.
    Mr. McHenry. OK. And the procedure with those that have 
fingerprints that cannot be read, will they be hired like they 
were hired in this address change?
    Dr. Groves. That is the standards we are after. I don't 
know how strongly I can say this, Congressman, but the safety 
of the U.S. public is of paramount interest to us, and I am 
committed to doing everything I can to achieve that.
    We have been working with the FBI after address canvassing 
and we have made various changes, and they are, under the 
guidance of the FBI--you know, what happens is that as you age, 
as all of us age, our fingerprints get harder to read.
    Mr. McHenry. True.
    Dr. Groves. The people who didn't have read fingerprints 
tend to be older and tend to be female, empirically.
    Mr. McHenry. Yes.
    Dr. Groves. So older women have harder problems in getting 
fingerprints read.
    Mr. McHenry. My time is limited. I know this is important, 
but if you could submit for this committee----
    Dr. Groves. I would be happy to give you a detailed list of 
things we are doing.
    Mr. McHenry. Because I think the GAO in their report would 
like to see that, as well.
    Dr. Groves. I would be happy to do so.
    Mr. McHenry. I have another additional thing. I read in USA 
Today yesterday that the expected response rate for mail, the 
initial form that will be mailed out on census day, is 64 
percent. This is, I think, new information. It was 67 percent 
in 2000. There have been some very substantial changes in that. 
We are re-mailing, in essence, those that do not respond via 
mail, which was not done in 2000, so the response rate was 
supposed to be better than it was in 2000 because of that 
procedure, alone. Why has this been reduced? The 3 percent 
would equal over $100 million by the initial cost estimates 
that we have, so it is real money we are talking about.
    Dr. Groves. Yes.
    Mr. McHenry. And I want to understand why this wasn't 
brought to our attention earlier and what your answer is on 
that, as well.
    Dr. Groves. Yes. First of all, the 64 percent number I 
didn't approve so I don't know where that number came from. We 
are actually estimating that number over and over again. That 
is a number that will be re-estimated over the coming months.
    Second, it is important and relevant for the committee to 
know that the response rates of every major national survey in 
the United States and every western country is declining. Those 
response rates are declining each year. The American community 
survey on the mail return rate is declining at between 0.5 and 
1 percentage point a year. We have a population that is tougher 
to measure than it was in 2000. We have, indeed, put in the 
design features you talked about that go in the other direction 
that should push it up, but the big changes in the population 
is a massive rock to push up the hill and we don't know yet how 
well these design features will work.
    Mr. McHenry. I mean, I read the press report that says 
Census Bureau analysis, which I'd like to know what report----
    Dr. Groves. Well, there are a lot of Census Bureau 
analyses.
    Mr. McHenry. Well, I'd like to see what you would submit as 
in what you think the result would be, because certainly with 
your history you certainly have specialized knowledge in this. 
We would like to have that just----
    Dr. Groves. I would be happy to.
    Mr. McHenry. OK.
    Thank you, Mr. Chairman.
    Mr. Clay. Thank you, Mr. McHenry.
    Dr. Groves, we know that the accuracy of the master address 
list relies heavily on close partnership between the Bureau and 
local governments. Looking to the future, this partnership will 
again be called upon year after year to help the Bureau produce 
accurate annual estimates. We know that the Bureau is altering 
the census challenge methodology. Will the census be working 
with local government officials like planners in developing the 
new methodology? And if so, how?
    Dr. Groves. I want to make sure I understand the question, 
Congressman. Are you talking about the population estimates 
program?
    Mr. Clay. Yes.
    Dr. Groves. OK. Well, as you know, the population estimates 
program is undergoing review. We have had outside experts in. I 
am very interested in this program, to improve it over the 
coming decade. We are going to have a lot more dialog about how 
best to do that. It is a set of technical issues about how, in 
the middle of the decade, you can get the best estimates.
    The procedures that have been used to work cooperatively 
with communities to update those estimates are worth reviewing. 
There are complaints kind of on both sides that ought to be 
aired, and I want that to happen, and I can assure you that 
will be an open process.
    Mr. Clay. Can the new construction program be expanded to 
include all additional addresses that cities might have missed 
in the LUCA process?
    Dr. Groves. Well, as you know, the new construction program 
is limited to those local governments that have access to new 
construction administrative records, and that is a smaller set 
than, for example, the State governments aren't used for that 
purpose.
    Right now under the current legislation we are limited to 
governmental entities providing us those updates, but the world 
is changing and it is worth talking about the future in various 
ways.
    Mr. Clay. You know, we are aware that the Census plans to 
hand-deliver 1.2 million questionnaires to residents in the 
Gulf region. This is a great start, but my concern is the 
followup. With the lack of mail receptacles and home telephone 
service in some of the affected areas, what additional measures 
will be taken for non-response followup?
    Dr. Groves. You know, Congressman, I just spent yesterday 
at our facility that is assembling the packets for the update. 
We call this update leave. It is really cool. You ought to 
visit it some time. But there are big plastic bags that contain 
little plastic bags that have questionnaires and a letter from 
me inside, so they are protected from the rain. They have 
little hooks on them so you can hook it inside a screen door if 
you don't have a mailbox of a house that is clearly 
inhabitable, thinking of the Gulf Coast areas you were just 
talking about.
    Mr. Clay. Sure.
    Dr. Groves. So I think we are thinking about the same 
things, and so far I think we are prepared for that. I can't 
wait to see how well that works. It is an area that is rapidly 
changing, as you know.
    Mr. Clay. I know on one of my field visits to the Gulf 
Coast region I was told by census workers that they, in some 
areas, they have to take boats into the bayou and other places 
in order to actually, I guess, verify addresses first, and now 
I guess they will have to drop those packages off by boat also.
    Dr. Groves. You would not believe the kind of 
transportation our enumerators are seeking. We also had a 
request for mules on some Indian reservations, because you 
literally cannot drive a four wheel drive vehicle up to some of 
the lodging.
    Mr. Clay. OK.
    Dr. Groves. So we do a lot of efforts that are unusual.
    Mr. Clay. Wow, you are really preparing for this, aren't 
you? Please detail the update and leave program that is 
utilized in rural and Gulf Coast areas. Upon recognition of the 
addresses of hidden housing units, will there be enough time to 
input these found addresses before non-response followup?
    Dr. Groves. I want to get your question right. Could you 
repeat that? I want to make sure I understand effectively.
    Mr. Clay. Detail the update leave program.
    Dr. Groves. OK.
    Mr. Clay. That is utilized in rural and Gulf regions.
    Dr. Groves. In the Gulf Coast areas, say take New Orleans, 
in Orleans, Planquemine, and St. Bernard Parishes, these were 
almost all mail-out/mail-back in 2000. With the collaboration 
of local government and civic leaders, we have identified all 
three of those parishes are going to be entirely update leave, 
so we will have people on the streets going structure-by-
structure. When a structure, according to a set of fixed rules, 
is defined as habitable, they will put a questionnaire on those 
structures.
    In areas of, say, Tammany Parish, the same thing will 
happen. So as you get away from the coast things get a little 
better, and there are certain areas that we will do update 
leave, but they may be surrounded by areas that are mail-out/
mail-back, and all of this is designed to be done in 
conjunction with local leaders who know what is happening.
    This is a rapidly changing thing. People are building 
houses now in New Orleans especially, and we have to be very 
current to get it right.
    Mr. Clay. And you did say St. Bernard Parish?
    Dr. Groves. St. Bernard's is fully update leave.
    Mr. Clay. Well, it just causes me to ask the question then: 
could this technique be applied in hard-to-count urban areas?
    Dr. Groves. It could, indeed, and a thing that is new---
actually this is relevant to Ms. Jacobs' testimony--one thing 
that is greatly expanded this decade I think is really 
something that is neat, and that is, for all of the census 
tracts--these are small geographical areas that were found to 
be hard to enumerate in 2000--there is a special plan for every 
tract. We have people who have already driven every street of 
those hard-to-enumerate tracts, and they have looked at every 
house on the tract, and they have asked the question how best 
to enumerate this area. If it is a mail-out/mail-back area that 
they are concerned about, they will do separate outreach 
efforts to encourage response. If there are other things going 
on, they have the freedom to tailor some of the methods they 
will use.
    I am very hopeful that this kind of customization down to 
the local level could pay off.
    Mr. Clay. That sounds impressive. Thank you for your 
response.
    We will now go to the gentleman from Georgia for 5 minutes, 
Mr. Westmoreland.
    Mr. Westmoreland. Thank you, Mr. Chairman.
    First question is for Ms. Jacobs. Ms. Jacobs, in your 
written testimony here I noted that I guess it was in 2003 your 
organization identified approximately 4,000 units that were not 
in the MAF, and this was before Mr. Groves, but with your 
working with the Census Bureau I think about 75 percent of them 
or a little over 3,000 of them were included in the MAF. Was 
this a satisfactory outcome for you?
    Ms. Jacobs. Thank you for the question, Mr. Westmoreland. 
We were, of course, thrilled with that outcome, but that 
outcome was, first of all, limited to the Los Angeles region 
and, second of all, it was based on a program that we undertook 
through our education and outreach program in 2000 and funded 
research. That was unusual, and it was, I would say, unique. 
And it has not been otherwise applied by the Census Bureau. We 
had census outreach workers who were hired by our organization 
identifying units that we thought had been missed or were 
highly likely to be missed, and it was only because the L.A. 
region was willing to cooperate in developing this address 
listing protocol with us and was willing to then take our 4,000 
units back, compare them to the master address file to which we 
had no access, of course, and then able to add units. But that 
was not something that had been necessarily approved by 
headquarters nor has it since been implemented by headquarters 
nor has it been approved or implemented in any other region, to 
our knowledge. But I think that is an approach that certainly 
could be used, and when I referred in my oral testimony to the 
address listing protocol that is what I meant.
    I think that the Bureau has implemented a similar protocol 
for the targeted non-sheltered outdoor locations, but that is 
going to be essentially a 1-day operation, and it is primarily 
to reach the homeless. It will not necessarily reach any of the 
hardest-to-count populations that we are concerned about.
    Mr. Westmoreland. I noticed that in Dr. Groves' testimony I 
believe he said that there was an outreach to approximately 
28,000 I guess different governments, whether it is city, 
county, State, or tribal, or whatever. Only about 14,000 of 
those had responded to, I believe, being in the program. Are 
you encouraging some of the local governments where you are at, 
because I know I read your protocol and what you are talking 
about, people being familiar with the area and the community 
being involved in this. Do you see a role for some of these 
governments to play in it that could help in that?
    Ms. Jacobs. Certainly yes, and I think that the Census 
Bureau has a good partnership program, and CRLA community 
outreach workers are a part of that program. We participate on 
complete count committees in many local jurisdictions, and we 
encourage local jurisdictions to participate.
    One of the limitations of LUCA, however, is that community-
based organizations cannot participate. I think it would be an 
enormous benefit to local governments, as well as to the Census 
Bureau, to make use of the knowledge of community-based 
organizations on the ground in those communities that could 
really provide assistance to very strapped local governments.
    Mr. Westmoreland. Thank you.
    Dr. Groves, it is good to see you again. We do appreciate 
your filling of your commitment to come in front of us often 
and let us ask you questions.
    One comment I will make, you know, Mr. McHenry mentioned 
the cost overruns. I will tell you that I have had several 
people across the country in the real estate business tell me 
that the Census Bureau in different locations was paying 
anywhere from $52 to $55 a square foot for office buildings, 
and at some point if you want to come by the office I will give 
you some of those locations, because right now typical office 
space is anywhere from $8 to $10 a foot, so you might want to 
check some of that out for your cost overrun.
    But, Dr. Groves, does the Census Bureau pay any outside 
groups to add addresses to this master address file? Do you 
have a contract with anybody, pay anybody to do that?
    Dr. Groves. To add addresses to the file?
    Mr. Westmoreland. Yes.
    Dr. Groves. The base of the master address file for this 
decade started with the 2000 census master address file. If you 
go back into the history of this, we assemble records in that 
decade from various commercial sources, but we have been 
updating that now, so to know the origin of actually every 
address in there is kind of tough at this point. But this 
operation that we have done over this decade has relied heavily 
on the Postal Service. I don't know how you count that in your 
thoughts, but that has been a chief updating source for us.
    Mr. Westmoreland. OK. And one final question. It will be a 
short one, I think. How do you think the budget overruns in 
conducting the address canvassing over the summer will impact 
your ability to effectively administer the 2010 census?
    Dr. Groves. Those budget overruns are intolerable to me, 
Congressman, and I believe, as I mentioned previously I think 
to the committee, that part of it was from a flaw in the cost 
modeling logic. That logic has been changed. Our big operation 
going forward, as you know, is the non-response followup 
operation. We are undergoing two independent cost modeling 
schemes. One has been partially completed, the other is going 
on now. I want to compare multiple ways of estimating the cost, 
because I think that is the way to protect your estimation.
    This is a very complicated process. I don't want to imply 
that it is easy to do. It is very important, though, to get 
this right.
    I am pleased that this operation that we are doing right 
now called group quarters validation, where we are going out to 
2 million addresses, appears to be on time, on budget, and that 
is a good thing. We can't tolerate these kind of overruns in 
our big operation, and it is not going to happen on my watch as 
long as I am in this position, anyway, I will tell you.
    Mr. Westmoreland. Thank you, sir.
    Mr. Clay. Thank you, Mr. Westmoreland.
    Mr. Zinser, have you had further results or reviews of the 
paper-based operations control system? If so, please detail 
some of your findings.
    Ms. Zinser. Sir, we don't have any further results from the 
review of the paper-based operations control system, but we 
include that system as a top risk to the census because of the 
late change to paper-based operations. There are a lot of 
changes that have to be done to that system, and that has to be 
a front and center focus item for the census, and we count that 
as a major factor in trying to determine whether or not the 
costs are going to be contained.
    Mr. Clay. I see. Dr. Groves, did you want to add anything?
    Dr. Groves. I would be happy to comment. I agree. It is 
deserving of scrutiny by my two colleagues here and me 
continuously.
    Mr. Clay. OK.
    Dr. Groves. We have a big test, I want to announce, coming 
up around Thanksgiving. It is a big deal for us. We have to hit 
that. We hope to break the system in Thanksgiving in order to 
make sure it is robust for the real use.
    Mr. Clay. Thank you for that response.
    Mr. Zinser, how can the non-response followup operations be 
made more efficient? Please respond to the question in terms of 
cost and effectiveness. You do have some history with which to 
evaluate this, I assume?
    Ms. Zinser. I didn't catch the last part of the question, 
sir.
    Mr. Clay. I'm asking if you have any history to evaluate 
the non-response followup, its effectiveness and cost 
efficiency.
    Ms. Zinser. Yes, sir, I think our plan for evaluating non-
response followup is similar to how we handled address 
canvassing is that we dispatched teams out to the field to 
actually observe the non-response followup operation, and by 
putting our folks out in the field and observing how the 
enumerators are operating, we hope to identify problems early, 
alert the Census Bureau to those problems, and then they make 
changes. We did that, for example, with address canvassing when 
we observed a number of listers in five different regions that 
we were in not following procedures. They were supposed to go 
up and knock on the door so they could get a good map spot with 
the hand-held computer. They weren't doing that. We alerted the 
Census Bureau and they took corrective action.
    I think that is pretty much our strategy for covering non-
response followup.
    Mr. Clay. To what degree will the accuracy of the master 
address file be affected by the Census Bureau's inability to 
track schedules, costs, and risk management activities of this 
endeavor? Do you have any figures for this?
    Ms. Zinser. I don't think I have any figures for that, sir, 
but the construction of the master address file, as we have all 
testified here this morning, is a key operation, and what our 
suggestion is is that they do some data analysis of the quality 
of the master address file right now to include using housing 
unit estimates and some administrative records to match against 
the master address file to try to target those areas where 
there might be problems with the quality of the file.
    Mr. Clay. What specific risk management activities are 
behind schedule with regard to the master address file?
    Ms. Zinser. Are behind schedule?
    Mr. Clay. Yes.
    Ms. Zinser. Well, the Census Bureau has identified probably 
somewhere in the area of 24 high-risk areas, and they are 
developing contingency plans for probably around 11 of those. 
One of those is called the Housing Unit Duplicates and Misses. 
It is a contingency plan that they are working on, but it is 
not completed, and I don't think they have any scheduled date 
for completing it, so I would list that as a key area to get 
some progress on.
    Mr. Clay. I see. Thank you for that response.
    Mr. Goldenkoff, is it true that the FBI has continued to 
express concerns regarding the Bureau's poor paper ink 
fingerprinting quality? Can the FBI guarantee a quick turn-
around of check results following the fingerprint submissions? 
And if the FBI cannot guarantee a quick turn-around, what is 
the Bureau's contingency plan?
    Mr. Goldenkoff. I'm not aware of any contingency plan that 
the Bureau would have if there is no quick turn-around. These 
operations are very short-lived and very often the people are 
hired and will be hired during non-response followup, during 
the operation or right before it is to begin, so it is just 
really a very short window that the Bureau would have in order 
to conduct these fingerprints and get the results back.
    Mr. Clay. How can best practices be utilized to ensure the 
Bureau provides a more reliable cost estimate for additional 
endeavors such as non-response followup, especially in light of 
the 25 percent over budget for address canvassing?
    Mr. Goldenkoff. Well, certainly the Bureau does need to 
rely on best practices and employ them. GAO has put out a guide 
to best practices for cost estimation. This has been a 
longstanding weakness with the Bureau. What we have seen is 
that the Bureau's cost estimates have lacked detailed 
documentation. The sources and assumptions that they have made 
were very weak or lacking. They were not comprehensive in the 
sense that all costs weren't included. And one of the things I 
would just like to bring up right now, you know, we have heard 
talk, Dr. Groves had mentioned about they are revisiting the 
mail response rate. Well, that has a huge impact for the final 
cost of the census. A 1-percentage point change in response 
rate, by the way, can have tens of millions of dollars worth of 
implication for the final cost of the census. So that would be 
something, right off the bat, that, you know, it is great that 
the Bureau is looking at that, but the question I would have is 
to what extent is that being reflected in their cost estimates.
    Mr. Clay. Thank you.
    Mr. Chaffetz, 5 minutes.
    Mr. Chaffetz. Thank you, Mr. Chairman.
    Again, thank you all. I know you all believe that the 
integrity in the process is paramount to successful completion 
and confidence by the American people.
    Director Groves, I would like you to get as specific as you 
can about the policy and procedures--hopefully they are 
written--as to the criteria what would exclude somebody from 
joining on and joining the census. We hear reports about 
criminals being hired to conduct the census. That is concerning 
to a lot of people, including myself. But what point, at what 
threshold do you say this person is not qualified, and to what 
degree are people qualified even though they might have a 
criminal background?
    Dr. Groves. I can supply that information. As you might 
guess, Congressman, publishing that information would provide a 
set of people information that would allow them to gain that 
system in a way that might be harmful to the safety of the U.S. 
public. I can tell you how we go about this.
    We receive from the FBI, on those names or fingerprints 
that generate a criminal history, the nature of the offenses. 
As you know, the FBI database doesn't completely give the 
disposition of all those offenses. So we review this, we give a 
chance to the applicant to provide counter information, and 
there are a set of crimes that are basically more serious than 
others where the applicant would fall out.
    I can tell you in the fingerprinting side that about 58 
percent, I believe, of those that had a criminal history come 
back from the FBI based on fingerprints were eliminated from 
the group; 42 percent stayed in because these were crimes that 
were judged not to threaten the safety of the U.S. public.
    Mr. Chaffetz. I mean, at least according to what I'm 
reading on page 13 of the GAO report, midway through the first 
complete paragraph, of the 1,800 workers with criminal records, 
approximately 750, or 42 percent, were terminated. So it would 
be the other way around; 58 percent were actually allowed to 
stay. The number 58 would be consistent. Why? How do we allow 
somebody with a criminal record to participate in the 
enumeration process of the U.S. census? I can't think of any 
threshold that I would have any confidence in allowing somebody 
to go knock on Grandma's door and invite themselves in to 
further discuss very pertinent personal information. I don't 
understand what threshold of criminal activity is acceptable by 
the census.
    Dr. Groves. I would be happy to go through this process if 
you would like.
    Mr. Chaffetz. Yes. That is why I am asking the question.
    Dr. Groves. The list of crimes that I talked about before 
are things that--the process by which we make these judgments 
is totally driven by our concerns about the safety of the U.S. 
public.
    Mr. Chaffetz. So why not exclude all of them?
    Dr. Groves. Because in the judgment of the process going 
on, these don't harm the safety of the American public. They 
can.
    Mr. Chaffetz. And that is a subjective point of view that 
you are just personally making on some--I mean, who is making 
these decisions?
    Dr. Groves. I would be happy to review this with you 
whenever you want, Congressman.
    Mr. Chaffetz. Right now would be ideal.
    Dr. Groves. I don't have the list of the offenses in front 
of me, but I could----
    Mr. Chaffetz. I am concerned that it is a subjective 
criteria; it is not an objective criteria. I think if you are 
going in and you are asking for personal, sensitive information 
about their names, their addresses, about what my 8-year-old 
daughter's birthday is, I can't find anybody with a criminal 
record that I would be comfortable giving that information to. 
Meanwhile, we have literally millions and millions and millions 
of good, hard-working, honest Americans without criminal 
backgrounds that are just dying to get an appointment. I can 
find no excuse for allowing somebody to deal with that 
sensitive information in the American people's homes.
    Based on the information I am seeing, Mr. Chairman, we have 
over 1,000 of them, at least, and that number is probably much, 
much greater than that. I have a deep, deep concern. GAO points 
out that crimes such as rape, manslaughter, child abuse are 
being dismissed. I appreciate that, but there are a whole lot 
of other crimes that I wouldn't express confidence in, either.
    Mr. Chairman, I see that my time has expired. For the 
record, I would like to exercise my right to have 5 minutes for 
each member of the panel, but I want to be sensitive.
    Mr. Clay. We will have another round of questions.
    Mr. Chaffetz. Fair enough. I yield back my time. Thank you.
    Mr. Clay. Thank you.
    Mr. Chaffetz. My expired time.
    Mr. Clay. Ms. Watson, you are recognized.
    Ms. Watson. Thank you, Mr. Chairman. I really want to thank 
you for having this hearing today. I can't think of a more 
significant and crucial hearing when it relates to the Census 
Bureau's ongoing compilation of the master address file. And, 
based on the time-tested theory that the quality of output can 
only be as good as the quality of input, the MAF is key to what 
we can expect to learn about the United States and the changes 
its population has gone in the last decade.
    I think everyone knows I come from the largest State in the 
Union, California, and 38 million people, and we get 2,000 
immigrating into our State on a daily basis, and they 
surprisingly don't all come from over the border, they come 
from across the Pacific, and so we are the first State in the 
Union that is a majority of minorities, and that brings a whole 
lot of different qualities to the count.
    I am sorry I missed the first part of the panel, because I 
know you have given us very vital information, and so I would 
like to know what challenges must be met to ensure a one-to-one 
match between the residents of the Nation and the Census 
Bureau's address book, and who we know is likely to be left out 
or under-counted, misunderstood, or intimidated into 
concealment if the Federal Government's message, methods, or 
motives lack transparency.
    Now, once every 10 years I bring someone from the census 
into my office, and I say, ``Let me tell you, if you are going 
to get an accurate count in my District you are going to go 
upstairs over the liquor store, you are going to come out on 
Sunday without your clip board and you are going to go to the 
playground at the school or at the church or at the park, 
because that is where you are going to find a lot of people 
coming out of those cramped apartments with their children.''
    So much depends on getting a picture of who we are in 
America and how we go about counting them. And so, if I can, 
Dr. Groves--and if you have already addressed this, then I will 
take the record as my information--but an amendment has been 
proposed in the Senate's appropriation bill for fiscal year 
2010 that would require the 2010 census to ask about 
citizenship, and how do you believe this would affect the 
Bureau's ability to perform a full and accurate count?
    Dr. Groves. Thank you for that question. As you may know, 
we have printed over 400 million forms already. I visited one 
of our facilities that had seven stacks of pallets in 400,000 
square feet filled with printed forms already stuffed, ready to 
be delivered. The most serious problem of changing the census 
now----
    Mr. McHenry. Dr. Groves, I think staff has a photograph of 
the very scene you are saying. I don't intend to take your 
time, and I ask, Mr. Chairman, if you would make her time 
whole, but there is a picture that you are referencing of those 
printed forms. I thought it would be a useful visual for those 
here today.
    Dr. Groves. It is an impressive picture, I believe.
    Mr. Clay. Go ahead. You may proceed, Ms. Watson.
    Ms. Watson. Dr. Groves was responding.
    Mr. Clay. OK.
    Dr. Groves. So the Census Bureau, before I got there, 
followed the regulations on how we should behave, and in 2007 
the topics of the questionnaire were submitted to the Hill for 
comment, for addition. That was a moment to add a topic. And 
then in the middle of 2008 we delivered to the Hill, according 
to regulations, the exact questions to be asked, and at that 
time there were no added questions asked.
    I can say with absolute confidence that if we add a 
question to this census questionnaire at this point we will not 
deliver the reapportionment counts on December 31, 2010. We 
will not provide the data for redistricting. We don't have 
enough time to make these changes.
    Ms. Watson. So what you are saying is that you have not 
considered immigration on this form?
    Dr. Groves. I am saying that the addition of a question 
about----
    Ms. Watson. You have not--I just want to know, is there any 
indication, any question relative to immigration on the form? 
Yes, no?
    Dr. Groves. Not at all. Not at all.
    Ms. Watson. OK. All right. Now, I am also concerned about 
home foreclosure and the number of people who have been forced 
out of their homes and on the streets, and the rising jobless 
rates means more Americans are leaving their homes and living 
in a constant shifting and non-traditional arrangements such as 
in their cars, in tent cities, and on the couches of various 
friends and family members, and all the while increased 
financial hardships may make some Americans less willing to 
cooperate with the census workers. What challenges has the 
economic and housing foreclosure crisis posed to the collection 
of a complete master address file?
    Dr. Groves. There are two things I am worried about. I am 
worried less about the master address file than the actual 
enumeration. The foreclosed homes--I was in L.A. 2 days ago--
are largely empty now. Some are not habitable now. We are going 
to spend money on those houses by knocking on the door after we 
mail a questionnaire to those houses, so there is a cost 
implication of foreclosures.
    I am also worried about the doubling up problem of homes 
that are the result of that foreclosure. There, we are 
redoubling our efforts to make sure people get the message if 
you have some family members living with you in that state, to 
include them where they live with you, even though you may not 
think of them as part of your home permanently. Because they 
don't have another residence, they need to be counted where 
they are.
    And then the other thing you mentioned is----
    Ms. Watson. What about those living in their cars?
    Dr. Groves. Yes. In L.A. I rode street after street where 
there are RVs parked one after the other and they stay there 
for 3 days and then they move to another neighborhood because 
of the parking regulations. There are people living in these 
RVs. These are people who were well off enough to have an RV 2 
years ago, but that is all they have now.
    This is a challenge for us in what Ms. Jacobs talked about, 
this 3-day period where we measure these non-traditional living 
situations. It is a new challenge for us. Our local regional 
folks are all over this problem, but it is going to be a 
challenge.
    Ms. Watson. I will agree with you. I represent Los Angeles, 
Culver City, Hollywood. I see them on the streets every day. An 
under-count has been constant in various areas. Every decade 
there is an under-count. Therefore, the representation is off a 
little bit, the resources that would flow in that would follow 
the numbers in various categories we lose. So it is really 
important.
    I think that even numerators who have been incarcerated can 
be rehabilitated and can be very helpful in some areas of the 
community where they recognize these people and they feel more 
comfortable giving up the information about how many live in a 
particular house and so on. I mean, I have apartments in my 
District where they hot bed. There might be a dozen people in a 
one-bedroom apartment. Somebody whose face might be familiar, 
somebody who has the charisma and so on, non-threatening type 
could probably give us a more accurate number. So I am all for 
your figuring out ways to count these people.
    I think my time might be up, but anyway let me go on to Mr. 
Zinser. If it is up, Mr. Chairman----
    Mr. Clay. It is up, and we will have a second round of 
questions.
    Ms. Watson. All right. I will yield back and I will wait.
    Mr. Clay. I thank the gentlewoman.
    Mr. McHenry, you are recognized for your second round of 
questions.
    Mr. McHenry. Thank you, Mr. Chairman.
    Mr. Goldenkoff, I want to start with you. I actually 
reviewed your Senate report and testimony, and you have added 
quite a bit to your testimony today, but the one essential part 
that is largely the same, I believe, is dealing with the cost; 
is that true?
    Mr. Goldenkoff. That is correct.
    Mr. McHenry. So forgive me if I am referencing the Senate 
section here, but you report, for example, the Bureau had 
planned for 25 percent of new hires to quit before, during, or 
soon after training; however, the national average was 16 
percent.
    Mr. Goldenkoff. Yes.
    Mr. McHenry. Bureau officials said that not having to 
replace listers with inexperienced staff accelerated the pace 
of operation. The way you write that, it is basically a report 
from the Bureau. Did you find that was true, based on your 
analysis?
    Mr. Goldenkoff. Well, we did not independently confirm the 
numbers, but based on our knowledge of the census and census 
operations we could see how that could be both a cost savings 
and also make the operation more efficient. Basically, among 
other things, it saves you training, for example. The people 
work all the way through. They are also more experienced, they 
know their jobs better, there is not that learning curve, so it 
would certainly make sense that would improve the pace of the 
operation.
    Mr. McHenry. OK. So it is kind of interesting that perhaps 
that lesson, alone, to know that because of perhaps the 
economic situation that we are in that people are sticking 
through the job, right?
    Mr. Goldenkoff. That is correct.
    Mr. McHenry. And the applicants are stronger to begin with.
    Mr. Goldenkoff. That is correct.
    Mr. McHenry. All right. So therefore we see perhaps a 
better result from our non-response followup coming up, based 
on that experience?
    Mr. Goldenkoff. It is possible. That is correct. You can 
make the argument that there is a better work force going into 
non-response followup in the sense that they have employment 
history, they have a work ethic. In the past the Census Bureau 
has relied on for the 2000 census a part-time part-time 
strategy, basically part-time employees, under-employed people. 
But to the extent that you have a very skilled work force now 
that is looking for work, those people tend to be better 
employees and more responsible.
    Mr. McHenry. Additionally, what was the key cost overrun? 
What was the largest failure of the Bureau with this cost 
overrun?
    Mr. Goldenkoff. There was actually several reasons for it.
    Mr. McHenry. I know in your report----
    Mr. Goldenkoff. A big reason was they under-estimated the 
workload, the address canvassing workload. There were I think 
around 11 million additional addresses that they hadn't counted 
on. Some of those came from LUCA. Some of those came from other 
sources. Each one of those addresses had to be verified in the 
field, and that is labor intensive and costly.
    Mr. McHenry. Sure. And you mentioned 11 million, which it 
says in your report that 11 million addresses were included 
that were not in their original 2009 budget; is that true?
    Mr. Goldenkoff. That is correct.
    Mr. McHenry. And that was one of the largest dollar amount 
increases?
    Mr. Goldenkoff. I believe so. That, and the fact that they 
hired more listers than they needed to because they didn't 
stick to their staffing model.
    Mr. McHenry. OK. Dr. Groves, in light of this, what are you 
doing to make sure that you don't have a massive cost overrun 
for all the processes that we have going forward?
    Dr. Groves. Yes.
    Mr. McHenry. I mean, a 25 percent cost overrun is 
extraordinary.
    Dr. Groves. I think you could classify the things we are 
doing under two categories. I view this as a combination of 
top-down cost modeling where you take the 2000 estimates, you 
update them by what has changed, and then you derive a new cost 
estimate, versus a bottom-up approach where you get the 
components of the activities, you cost each one, and you 
aggregate it up.
    The typical Bureau approach is to do a top-down cost model. 
Those have been updated based on address canvassing results, 
the new hiring and attrition rates, and we have changed the 
staffing model going forward, so it took advantage of the 
information properly done. That is a good thing to do, but I 
don't think it is sufficient, and so we are also building a 
model from the bottom up, getting activity level cost and then 
aggregating it up, and I want to compare those two aggregate 
cost estimates.
    Mr. McHenry. My time has expired, but in closing, you know, 
you were here when address canvassing had ended. That was about 
the approximate time you testified, and you didn't want to 
discuss the cost of this in that hearing. With the 25 percent 
cost overrun, it seems to me unfathomable that you did not know 
that there would be significant cost overruns. What I have said 
in every meeting with you personally, in public, and what the 
chairman has said as well, and I think just about every member 
of this committee, is that we want to be of assistance here.
    If you keep us in the dark about challenges or problems, 
you know, including $88 million that was not budgeted for, it 
seems to me that you were not keeping us apprised of this, and 
that is rather disappointing, and I would hope that whoever is 
counseling you to hold back on that information, that you don't 
listen to that counsel; that you come forward and let us know 
as soon as problems occur, because we do want to be of 
assistance. We want to make sure that everything is there for 
you so you can have the best, most accurate counts. I know that 
is what the Bureau wants. I know that is what you personally 
want. But you need to keep us informed on this.
    I can understand if you didn't know a 3 percent cost 
overrun, but 25 percent, for heaven's sake, that seems to me 
unfathomable that you didn't know that. And so I would 
encourage you to come forward as soon as you know there are any 
problems or challenges, and we do want to be of assistance.
    Thank you.
    Mr. Clay. Thank you, Mr. McHenry.
    Ms. Jacobs, you mentioned in your testimony that many 
housing units of migrant and seasonal farm workers are not in 
the master address file because they are actively concealed. 
What do you mean by actively concealed?
    Ms. Jacobs. Well, I think there are several reasons for 
active concealment. What I mean about that is that, ranging 
from the owners and operators of that housing not wanting it to 
be seen and therefore disguising it or hiding it as best they 
can, to the occupants of that housing not wanting to be 
discovered because they are living in what would be considered 
illegal units or living where they think they are not supposed 
to. For example, they are hidden under tarps and lean-tos in 
the canyons in San Diego County, or in L.A. County in onion 
fields. They are living in between trees, under tarps. These 
are circumstances where they are trying not to be seen because 
they will be dislocated from the housing. They might be 
evicted. They might be threatened if they remain there.
    When the owners of the property or the owners of the 
illegal housing units believe that they will be either 
prosecuted or sued, for example, by CRLA for maintaining 
uninhabitable dwellings for their workers, they will try to 
actively conceal those units, so that can be difficult for the 
Census Bureau. But I don't think active concealment is the 
biggest problem that we have.
    Mr. Clay. Well, how can the Census Bureau get a better 
count of these populations?
    Ms. Jacobs. Well, I think the example that we gave in our 
written testimony, which was also cited in 2003 by the GAO 
report on the address listing protocol that we used, is the 
best example. We have trusted faces in the community. We have 
reliable community outreach workers that work for local 
community-based organizations who know where this housing is 
located who can work in partnership with the Census Bureau to 
assist both in address listing and in enumeration of these 
locations. They know where they are. They are trusted by the 
occupants. They can go a long way to opening doors, so to 
speak, for the Census Bureau.
    Mr. Clay. You know, the deplorable conditions for housing 
for immigrant workers are not confined to migrant farm workers. 
Recent immigrants to this country have the same living 
conditions in cities, just as Ms. Watson pointed out. What are 
some examples of low-visibility units in cities that do not 
have postal addresses, and how are they reached by Census 
workers?
    Ms. Jacobs. I think that the problems in isolated rural 
areas and the types of housing in which migrant and seasonal 
farm workers live are very similar to the small towns in 
agricultural communities, as well as many of the inner city 
areas where there are illegal units that are being rented out, 
there are illegal garages that are being rented out. Underneath 
someone's porch is being rented out as a ``habitable'' 
dwelling. Again, there is local knowledge of where these units 
are and, more importantly, I think that the Census Bureau's own 
hard-to-count database can be improved, enhanced, and utilized 
to target special enumeration procedures in areas that have a 
high concentration of hidden housing units. I think that can be 
used in rural areas, as well as in inner city urban areas, and 
it should be done regardless of whether those areas are 
considered mail-out/mail-back areas. They still need to be 
targeted for enumeration that is not done by mail, or we will 
miss not only entire housing units but we will continue to have 
people omitted from households.
    Mr. Clay. You know, speaking of omissions, in 1990 it was 
estimated that 48 to 52 percent of the migrant seasonal farm 
workers were under-counted. A large part of the under-count was 
attributed to total household omissions. What is the extent of 
these problems heading into the 2010 decennial? What do you see 
them as?
    Ms. Jacobs. I believe that 2010 will have very similar 
problems. I give the Census Bureau credit for developing 
improved job aids and improved instructions and training to 
their address listers and enumerators, but I think that a lot 
more needs to be done in order to ensure that the locations are 
identified for the Census Bureau so that addresses can be added 
to the address file at any time through non-response followup, 
as well as during the decade, and so that again these areas can 
be targeted for special enumeration procedures.
    Mr. Clay. What suggestion do you have for collaboration 
between the Bureau and groups like yours to get these addresses 
in the file?
    Ms. Jacobs. Well, we certainly take advantage of the 
partnership opportunities that the Census Bureau offers and we 
encourage and we train other community-based organizations to 
do the same. I think, however, that the Census Bureau could 
make better and more use of local knowledge and community-based 
organizations in its LUCA process, as well as in address 
canvassing and by using the special protocol that we described 
in our written testimony.
    Mr. Clay. Thank you so much for your response.
    The gentleman from Utah is recognized.
    Mr. Chaffetz. Thank you, Mr. Chairman.
    Mr. Goldenkoff, I would like to go back to the questioning 
we were doing with Director Groves, talking specifically about 
the criminal element that seems to be acceptable to Census 
Bureau. Can you give me your perspective on that situation? Are 
there criteria? Are they objective? Subjective? How many people 
are we talking about here?
    Mr. Goldenkoff. I don't know. We have not seen the actual 
list. I mean, obviously some crimes are less severe than 
others, but which ones would allow you census jobs and which 
ones will not, we don't know because we haven't seen the list.
    Mr. Chaffetz. Is there any written criteria for this, or is 
this just something that is just done on the fly and very 
subjective?
    Mr. Goldenkoff. We have not looked into that, so I do not 
know.
    Ms. Zinser. Sir, can I address that?
    Mr. Chaffetz. Yes. Yes, please.
    Ms. Zinser. I became the IG at Commerce in late December 
2007/January 2008. By Federal of 2008 we were alerting the 
Department and the Census Bureau that they had to get on this 
fingerprint issue.
    Mr. Chaffetz. Yes.
    Ms. Zinser. And they weren't prepared for it, and it took 
them 6 to 8 months to get ready for fingerprinting, and the 
original estimates for fingerprinting were up in the $600 
million and $700 million. Our office worked with them for 6 or 
7 months, working on their cost estimate.
    There is criteria, and there is criteria in other Federal 
jobs, also. My most recent experience before Commerce was at 
transportation and transportation security. There are common 
lists of offenses that the Federal Government refers to in 
terms of whether somebody is qualified or disqualified. One of 
the things we recommended for the Census Bureau was, in the 
past they would let local or regional offices make 
determinations on which crimes are disqualifying and which 
crimes aren't. We recommended that they centralize that in an 
office called CHECK. I can't tell you what that exactly stands 
for, but there is an office in Census headquarters that has 
centralized these kind of determinations, and we think that is 
a good practice.
    Mr. Chaffetz. Is there written criteria?
    Ms. Zinser. Yes, there is. I think that doesn't eliminate 
all subjectivity. I think you do have to make some judgments. 
For example, how long ago the offense occurred? Is it a 
misdemeanor? Is it a felony? Is it a violent crime, non-violent 
crime?
    And I also know that the Census Bureau has been consulting 
with the FBI on those types of issues.
    Mr. Chaffetz. My concern is that we just are somewhat in 
the dark on this. Again, it is giving the confidence of the 
integrity of the process and the people that are going to be 
knocking on their door, because, unlike most other Federal 
jobs, they are actually going to be going up and approaching 
somebody in their home and asking for sensitive information 
that can lead to other nefarious types of activities, and thus 
the concern.
    There are certainly a number of other types of Government 
jobs that somebody with a criminal background can participate 
in, whether it is the Department of Transportation or Member of 
Congress, whatever it might be, right? But in terms of the 
census, I think there is a great deal of sensitivity.
    Going back to Director Groves, how many people are we 
talking about, because, at least the way I read and interpret 
the numbers from the GAO report, we are talking over a thousand 
people. It is not a small, ``Hey, we have a handful here or 
there.'' I recognize the totality of the effort that is going 
on, but this seems like a rather large--and I sense a degree of 
secrecy that you want to keep from this committee in allowing 
us to understand so they can't ``game the system'' that I just 
find wholly unacceptable. I think there is a great deal of fear 
that will be created probably on the other end of it by being 
so secretive about what is acceptable and what is not 
acceptable.
    Dr. Groves. I can say, Congressman, that everything we do 
is in compliance with OPM guidelines that are, indeed, 
published. I don't have those with me, but I can supply those. 
So we are following that as well as we can.
    I think the other thing to note, just to make sure that I 
am communicating the facts correctly, is that the existence of 
a record in the FBI doesn't imply conviction of a crime.
    Mr. Chaffetz. So if somebody has been charged with a crime 
but not convicted, are they allowed to be an enumerator?
    Dr. Groves. What happens then is that the applicant is 
required by us to provide court certified documentation on the 
outcome of the case.
    Mr. Chaffetz. If they have----
    Dr. Groves. That is following the OPM guidelines.
    Mr. Chaffetz. And we are talking about tens of thousands of 
people here who have not completed the background process; that 
is, they have not had their fingerprinting processed by the 
FBI. Are those people allowed to start work, even though they 
haven't completed that process? Because it looks like, based on 
what has been going on in surveying and going out to all the 
neighborhoods and trying to figure out the maps and all that, 
that those people have actually been employed and working, 
despite what ultimately concluded was unacceptable, nefarious 
behavior.
    Dr. Groves. This group has universally passed the FBI name 
check that is based on name, date of birth, Social Security, 
and found----
    Mr. Chaffetz. Do you use e-verify?
    Dr. Groves. We do use e-verify as part of the employment 
process. In addition to that, then we do the FBI name check. So 
everyone has passed that.
    Mr. Chaffetz. So all of those names have gone through the 
e-verify process?
    Dr. Groves. All of those names, to my knowledge, have gone 
through the e-verify process.
    Mr. Clay. The gentleman from Utah's time has expired.
    Mr. Chaffetz. Understood. Thank you, Mr. Chairman.
    Mr. Clay. The gentleman from North Carolina?
    Mr. McHenry. Thank you, Mr. Chairman. Just briefly.
    Dr. Groves, does the Bureau have a set of internal 
procedures and policies on what is a disqualifier in terms of 
criminal records?
    Dr. Groves. Yes.
    Mr. McHenry. OK. Mr. Goldenkoff, has that been verified by 
GAO?
    Mr. Goldenkoff. Yes.
    Mr. McHenry. Yes.
    Mr. Goldenkoff. They do.
    Mr. McHenry. OK. Mr. Zinser.
    Ms. Zinser. Yes, sir, I have seen them.
    Mr. McHenry. All right. Mr. Groves, are you confident that 
there are no violent criminals that work for the Census Bureau?
    Dr. Groves. I am confident that the people employed by the 
Census Bureau have gone through this process and have been 
judged as not having a criminal history under the process.
    Mr. McHenry. Sure. And would the three of you agree to 
followup with Mr. Chaffetz and his staff in regards to his line 
of questioning?
    Dr. Groves. I would be happy to.
    Mr. Goldenkoff. Yes, sir.
    Ms. Zinser. Yes.
    Mr. McHenry. OK. Thank you all for that, and thank you for 
your testimony today.
    Mr. Chaffetz. Would the gentleman yield?
    Mr. McHenry. Sure.
    Mr. Chaffetz. Thank you. I won't keep everybody all day, I 
promise.
    Mr. Clay. The gentleman is recognized.
    Mr. McHenry. I asked the chairman for 1 minute, so I think 
you have a----
    Mr. Chaffetz. OK. Thank you.
    What percentage of the people going through the process go 
through the fingerprinting process? It is 100 percent; is that 
correct?
    Dr. Groves. We don't hire anyone who doesn't pass the name 
check. All the people who pass the name check then are 
submitted to the fingerprinting--who we wish to hire, are 
submitted.
    Mr. Chaffetz. And how long does it take? You know, they 
fill out their application, and then--I mean, the FBI, are they 
taking----
    Dr. Groves. This is done on the first day of training. 
There are two cards made.
    Mr. Chaffetz. Right.
    Dr. Groves. By two different fingerprinters. Those cards 
are FedExed to our national processing center and then 
electronically transmitted to the FBI. The turn-around time on 
the FBI in the last operation was about 22 hours. That process 
seems to be working. We beefed up the electronic pipeline to 
that, and we are doing a big load test of that. We are going to 
simulate a million hires through the FBI's submission process, 
just to make sure we can do that volume when we have to.
    Mr. Chaffetz. What is the biggest concern that you have at 
this moment? All things considered, what is your biggest 
concern?
    Dr. Groves. About what? About the entire 2010 census?
    Mr. Chaffetz. The entire totality of the process. When you 
wake up in the morning and say, ``Oh, my goodness, this is my 
biggest concern,'' what would it be?
    Dr. Groves. I'm most worried about the behavior of the 
American public, whether they will return this questionnaire at 
the rates we hope they will, and that the leadership of this 
country ignites and energizes themselves to encourage that 
participation. We need you at this moment.
    Mr. Chaffetz. Thank you.
    Thank you, Mr. Chairman. I appreciate your indulgence.
    Mr. Clay. Thank you, Mr. Chaffetz.
    The Director couldn't summarize it better. I want to thank 
the panel for your testimony today.
    That concludes the hearing. Hearing adjourned.
    [Whereupon, at 4:17 p.m., the subcommittee was adjourned.]