[House Hearing, 106 Congress]
[From the U.S. Government Publishing Office]




 OVERSIGHT OF THE 2000 CENSUS: MAIL-BACK RESPONSE RATES AND STATUS OF 
                             KEY OPERATIONS

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

                                HEARING

                               before the

                       SUBCOMMITTEE ON THE CENSUS

                                 of the

                              COMMITTEE ON
                           GOVERNMENT REFORM

                        HOUSE OF REPRESENTATIVES

                       ONE HUNDRED SIXTH CONGRESS

                             SECOND SESSION

                               __________

                             APRIL 5, 2000

                               __________

                           Serial No. 106-186

                               __________

       Printed for the use of the Committee on Government Reform


  Available via the World Wide Web: http://www.gpo.gov/congress/house
                      http://www.house.gov/reform

                               __________

                    U.S. GOVERNMENT PRINTING OFFICE
70-056                     WASHINGTON : 2001


_______________________________________________________________________
 For sale by the Superintendent of Documents, U.S. Government Printing 
                                 Office
Internet: bookstore.gpo.gov  Phone: (202) 512-1800  Fax: (202) 512-2250
               Mail: Stop SSOP, Washington, DC 20402-0001


                     COMMITTEE ON GOVERNMENT REFORM

                     DAN BURTON, Indiana, Chairman
BENJAMIN A. GILMAN, New York         HENRY A. WAXMAN, California
CONSTANCE A. MORELLA, Maryland       TOM LANTOS, California
CHRISTOPHER SHAYS, Connecticut       ROBERT E. WISE, Jr., West Virginia
ILEANA ROS-LEHTINEN, Florida         MAJOR R. OWENS, New York
JOHN M. McHUGH, New York             EDOLPHUS TOWNS, New York
STEPHEN HORN, California             PAUL E. KANJORSKI, Pennsylvania
JOHN L. MICA, Florida                PATSY T. MINK, Hawaii
THOMAS M. DAVIS, Virginia            CAROLYN B. MALONEY, New York
DAVID M. McINTOSH, Indiana           ELEANOR HOLMES NORTON, Washington, 
MARK E. SOUDER, Indiana                  DC
JOE SCARBOROUGH, Florida             CHAKA FATTAH, Pennsylvania
STEVEN C. LaTOURETTE, Ohio           ELIJAH E. CUMMINGS, Maryland
MARSHALL ``MARK'' SANFORD, South     DENNIS J. KUCINICH, Ohio
    Carolina                         ROD R. BLAGOJEVICH, Illinois
BOB BARR, Georgia                    DANNY K. DAVIS, Illinois
DAN MILLER, Florida                  JOHN F. TIERNEY, Massachusetts
ASA HUTCHINSON, Arkansas             JIM TURNER, Texas
LEE TERRY, Nebraska                  THOMAS H. ALLEN, Maine
JUDY BIGGERT, Illinois               HAROLD E. FORD, Jr., Tennessee
GREG WALDEN, Oregon                  JANICE D. SCHAKOWSKY, Illinois
DOUG OSE, California                             ------
PAUL RYAN, Wisconsin                 BERNARD SANDERS, Vermont 
HELEN CHENOWETH-HAGE, Idaho              (Independent)
DAVID VITTER, Louisiana


                      Kevin Binger, Staff Director
                 Daniel R. Moll, Deputy Staff Director
           David A. Kass, Deputy Counsel and Parliamentarian
                    Lisa Smith Arafune, Chief Clerk
                 Phil Schiliro, Minority Staff Director

                       Subcommittee on the Census

                     DAN MILLER, Florida, Chairman
THOMAS M. DAVIS, Virginia            CAROLYN B. MALONEY, New York
PAUL RYAN, Wisconsin                 DANNY K. DAVIS, Illinois
MARK E. SOUDER, Indiana              HAROLD E. FORD, Jr., Tennessee
------ ------

                               Ex Officio

DAN BURTON, Indiana                  HENRY A. WAXMAN, California
                       Jane Cobb, Staff Director
              Lara Chamberlain, Professional Staff Member
                 Amy Althoff, Professional Staff Member
                        Andrew Kavaliunas, Clerk
          Cedric Hendricks, Minority Professional Staff Member


                            C O N T E N T S

                              ----------                              
                                                                   Page
Hearing held on April 5, 2000....................................     1
Statement of:
    Mihm, J. Christopher, Acting Associate Director, Federal 
      Management and Workforce Issues, U.S. General Accounting 
      Office, accompanied by Robert N. Goldenkoff and Mark Bird, 
      U.S. General Accounting Office.............................    58
    Prewitt, Kenneth, Director, U.S. Bureau of the Census, 
      accompanied by John Thompson, Marvin Raines, and Bill 
      Barron, U.S. Bureau of the Census..........................    28
Letters, statements, etc., submitted for the record by:
    Maloney, Hon. Carolyn B., a Representative in Congress from 
      the State of New York, prepared statement of...............    10
    Mihm, J. Christopher, Acting Associate Director, Federal 
      Management and Workforce Issues, U.S. General Accounting 
      Office:
        Letter dated June 7, 2000................................    81
        Prepared statement of....................................    61
    Miller, Hon. Dan, a Representative in Congress from the State 
      of Florida, prepared statement of..........................     5
    Prewitt, Kenneth, Director, U.S. Bureau of the Census, 
      prepared statement of......................................    33

 
 OVERSIGHT OF THE 2000 CENSUS: MAIL-BACK RESPONSE RATES AND STATUS OF 
                             KEY OPERATIONS

                              ----------                              


                        WEDNESDAY, APRIL 5, 2000

                  House of Representatives,
                        Subcommittee on the Census,
                            Committee on Government Reform,
                                                    Washington, DC.
    The subcommittee met, pursuant to notice, at 2 p.m., in 
room 2247, Rayburn House Office Building, Hon. Dan Miller 
(chairman of the subcommittee) presiding.
    Present: Representatives Miller, Ryan, Maloney, and Davis 
of Illinois.
    Staff present: Jane Cobb, staff director; Chip Walker, 
communications director; Lara Chamberlain and Amy Althoff, 
professional staff members; Andrew Kavaliunas, clerk; Michelle 
Ash, minority counsel; David McMillen and Mark Stephenson, 
minority professional staff members; and Jean Gosa and Earley 
Green, minority assistant clerks.
    Mr. Miller. Good afternoon. A quorum being present, the 
subcommittee will come to order. There will be a vote in a 
short period of time, but at least we can get started with our 
opening statements.
    Today we continue our series of oversight hearings into the 
2000 census. Coming before the subcommittee today will be Dr. 
Kenneth Prewitt, Director of the Bureau of the Census, and 
Christopher Mihm, Acting Associate Director, Federal Management 
and Workforce Issues, U.S. General Accounting Office.
    Before I go further, I would like to say to everyone 
listening or watching this hearing that if you haven't mailed 
in your census form, long or short, please take the time to 
fill it out and mail it back. The census can't be a success 
without your participation. The money needed to ensure that you 
have the roads, emergency services, day care, schools and other 
vital services are tied directly to the responses you give on 
your census questionnaire.
    If you don't have a questionnaire or are concerned that you 
might be missed, you can call the Census Bureau's telephone 
questionnaire assistance line for help. That number is 1-800-
471-9424. Let me repeat that, 1-800-471-9424.
    If you have already mailed in your form, thank you for 
doing your part to ensure that America is accurately counted.
    I've read Director Prewitt's testimony, and I must say that 
I am very impressed by the complexity of the current ongoing 
operations. For example, the Bureau deserves praise for the 
mail response Web site now available at www.census.gov. The 
ability for virtually any city or county to look and see their 
response rates daily, what it was in 1990, and how it compares 
to the national average is an important addition to this 
census.
    Today, there are a number of different issues that I would 
like to address: The ongoing recruiting efforts as we approach 
the most difficult stage of the full enumeration, the 
nonresponse followup, which will be the most demanding task 
facing the Bureau in the full enumeration; the current mail 
response rate, on which the success of the census hangs; and 
then the ongoing controversy regarding the long form 
questionnaire.
    Clearly the biggest controversy surrounding the census has 
been the perceived intrusiveness and the invasion of privacy of 
the long form. In 1998, the Census Bureau distributed this 
binder with the long form questions and explanations to all 
Members of Congress and the Senate and asked for comments. Few 
comments were received. Clearly, Members did not know at that 
time what the level of dissatisfaction would be just a mere 2 
years later.
    However, from the moment census forms were being received, 
it was clear that this was the No. 1 complaint received by the 
subcommittee. While the long form has always been less popular 
than the short form, the attitudes toward the 2000 long form 
seem to be particularly intense despite the fact that it is the 
shortest ever and only differs by one new question from 1990. 
During the 1998 dress rehearsals, the long form response rate 
was between 10 and 15 percentage points lower than the short 
form. However, this information was not provided to the 
Congress until June 1999, after the questionnaire had been 
approved.
    From the first day that the forms were being received at 
millions of homes around the Nation, Members of Congress were 
receiving phone calls from constituents who were very upset 
about the long form. While some in Congress tried to downplay 
the extent of the problem, it was clear to me that this would 
be the biggest issue next to sampling that we would have to 
deal with in this census.
    Every major newspaper in the Nation has written about the 
long form and the privacy issue. Electronic media from talk 
radio to television have weighed in. It would be a mistake or a 
callous political move to lay the blame for this controversy at 
the feet of Republicans. This Republican Congress has been 
nothing but committed to the census. Republicans have said from 
the start that the Census Bureau would get the resources it 
needed to conduct a fair and accurate census. Republicans have 
kept that promise. In fact, numerous Members have promoted the 
census in their districts in a number of different ways, 
including Census in the Schools events and public service 
announcements like the sample you will see now.
    [Videotape played.]
    Mr. Miller. The reason why there is a long form controversy 
is because millions of Americans aren't comfortable answering 
the questions, and while some are quick to wag their political 
finger, more thoughtful consideration on this topic will be 
more constructive. Long before remarks by any congressional 
leaders, news stories were talking about the long form 
problems. The News Hour on PBS had an entire segment on the 
privacy issue and the long form almost 2 weeks ago. On 60 
Minutes, one of the most popular news shows on television with 
almost 13 million viewers weekly, commentator Andy Rooney 
voiced to the Nation two Sundays ago his criticism of the long 
form. He concluded his commentary by saying, ``I am not going 
to fill out the long form. I'll send them about what a soldier 
has to give if he's captured in a war: my name, address and 
Social Security number. Otherwise, Census Bureau, count me 
out.''
    In my hometown in Bradenton, FL, my wife and I live next to 
an elderly woman in her eighties. She has trouble with her 
eyesight, so my wife assisted her in filling out her census 
form. There were several questions that she simply would not 
answer, including giving her phone number. She noted to my wife 
that Florida was a State that at one time sold its driver's 
license list, and she simply was not going to give her phone 
number to the Federal Government. And while we all know that 
the census operates in a confidential environment, I believe we 
must all realize that it is exceptionally difficult for 
government to separate its entities. A violation of privacy on 
the State or local level, in people's minds, translates to all 
levels of government, including the Federal level. To the 
average person, government is government.
    Another factor at work here is computer technology and the 
Internet age. While both have brought tremendous convenience to 
our lives, grown our economy and fundamentally changed the way 
Americans live, they each have also brought new privacy 
concerns. While our government reaps the benefits of our 
technological prosperity, government must also share the burden 
of new privacy concerns. I also believe, sadly, that some of 
the recent scandals involving this administration, particularly 
the misuse of the FBI files, have not helped in building 
America's trust in her government. And while no single cause 
may be blamed, clearly there has been a change in attitudes 
toward trust in government since the 1990 census. 
Unfortunately, the 2000 census is feeling some of the brunt of 
this distrust.
    So what does this all mean? What should people do who have 
that long form sitting on their coffee table or kitchen 
counter? To put it simply, fill it out and mail it in. Congress 
has heard the dissatisfaction with the long form loud and 
clear. However, to change our approach in the middle of the 
census is impossible.
    In the coming months, my committee will hold hearings on 
the long form and privacy issues. All sides will have an 
opportunity to come to the table and be heard. This includes 
privacy advocates who believe the information is not needed and 
government data users who say the information is indispensable.
    I must say, however, that this Congress will look to 
eliminate the long form for the 2010 census. Of course, we 
can't eliminate the long form in a vacuum. There is information 
that government needs to make informed decisions on the 
allocation of resources and the planning and distribution of 
$185 billion in funding. A new tool called the American 
community survey is being developed by the Census Bureau. Is 
that the answer? Maybe. This is going to take careful 
consideration by this subcommittee and eventually the Congress 
as a whole.
    What is clear is that Republicans and Democrats must both 
work to promote the census. If one side or the other attempts 
to gain political advantage over the other during these 
critical weeks, then surely participation in the census will be 
hurt. An inaccurate census hurts America. An accurate census is 
in everyone's best interest. This is your future. Don't leave 
it blank.
    [The prepared statement of Hon. Dan Miller follows:]

    [GRAPHIC] [TIFF OMITTED] T0056.001
    
    [GRAPHIC] [TIFF OMITTED] T0056.002
    
    [GRAPHIC] [TIFF OMITTED] T0056.003
    
    Mr. Miller. Mrs. Maloney.
    Mrs. Maloney. Thank you, Mr. Chairman. And welcome to our 
witnesses, Dr. Prewitt from the Census Bureau and Mr. Mihm from 
GAO. I've seen so much of you lately, it seems like we are 
becoming very old friends.
    April 1, census day, was 4 days ago, and major census 
operations are now under way. Though the most labor-intensive 
activities are yet to come, all signs now are good. The largest 
peacetime mobilization in our history is under way, and I 
salute Director Prewitt and the census staff for an excellent 
job to date.
    Right now, the key success indicator for the census is the 
mail-back response rate, how many households have mailed back 
their forms. As of today, that stands at 55 percent, or about 
67 million households. That still leaves 45 percent of our 
Nation's households that have not returned their forms, and I 
urge everyone who has not mailed their form back to do so 
today, right now.
    At 55 percent, however, it seems that the estimated 
response rate of 61 percent will be met, and I'm hopeful it 
might be exceeded. The Director has challenged the Nation to 
reach 70 percent, and I hope and think we might reach that 
mark. I don't want to sound too optimistic, but the hard work 
on the advertising campaign, the partnerships, and promotional 
activities appears to be paying off.
    Other indicators are positive as well. Recruiting continues 
to go well, with the Bureau reaching its goal of 2.4 million 
qualified applicants by March 31, almost 3 weeks ahead of 
schedule. 25.5 million forms have already been scanned with 
continued high accuracy.
    Update/leave operations were successfully completed on 
schedule, almost 6 million phone calls have gone to the 800 
number, and 58,000 forms have been completed on the Internet.
    The other night, I went out with Chairman Miller at 4 a.m. 
to watch the temporary employees that the census has hired from 
the community to count the homeless. It was incredibly 
impressive to see the dedication and commitment of this work 
force operating in the middle of the night in difficult and 
often hazardous areas. So, things are going about as well as 
could be expected operationally.
    Considering the doom and gloom of just a few months ago on 
both the hiring needs and the mail response rate, things are, 
in fact, going remarkably well. The two major concerns raised 
by the GAO last December, hiring and response rates, are 
clearly on track, which makes the recent comments about the 
long form by senior Republicans all the more unfortunate.
    Clearly one contingency that GAO could not warn us about 
are some of the irresponsible remarks that have been in the 
news lately by elected officials who should know better. Let me 
make clear I am not referring to the chairman of this 
subcommittee. He has been a supporter of the census and the 
long form throughout this latest turmoil. But several prominent 
Republicans, including Senator Lott, Governor Bush of Texas and 
J.C. Watts, Chair of the Republican Conference, have recently 
complained that the long form is too nosy, that it asks too 
many questions. Some of these individuals have even made public 
statements suggesting that Americans should not complete their 
forms, despite the fact that refusing to complete these forms 
would be a violation of Federal law.
    I think these comments are outrageous, irresponsible, 
pandering to fringe groups and the radio talk show circuit. 
They threaten the success of the census by driving response 
down.
    We have Members of Congress saying that they ``believe in 
voluntarily cooperating'' with the government, but beyond that 
they won't follow the law. Since when did following the law in 
this country become a voluntary thing? What is really 
disingenuous is the fact that most of the questions on the long 
form have been around for decades. In fact, Ronald Reagan 
signed off on every single question in the 2000 census during 
preparations for the 1990 census, except for one required this 
decade by welfare reform.
    Over 2 years ago, as the content of the long and short 
forms was being finalized, every Member of Congress received 
this book, a detailed list of the questions to be asked, 
including a description of the need for asking it, along with 
the specific legal requirements supporting it.
    So this controversy, at this late date, strikes some as 
intentional sabotage. At the very least it is willful disregard 
for a successful census. While it may not be intentional, it 
clearly shows an ignorance of how incredibly useful census data 
is, and how much of a difference it makes in the lives of 
millions of Americans.
    Let's look at the plumbing question the talk radio shows 
seem to focus on. Well, it may shock some, but there are places 
in this country where Americans don't have plumbing, in the 
Colonias in Texas, on Indian reservations, and I daresay 
probably in rural communities in Mississippi.
    Or let's look at question 17 concerning a person's 
physical, mental or emotional condition in the last 6 months. 
Are some Members saying they don't want to know how big a 
problem this is, how many disabled Americans there are in this 
country, how many disabled vets, and where there are high 
concentrations of them who need services?
    It is my understanding that some of these leaders have 
started to moderate their comments. Well, they shouldn't just 
moderate their comments, they should be in the forefront of 
urging all Americans to fill out their forms completely. They 
should be urging their members to join them in supporting the 
census, all of the census. Anything less is unacceptable. 
Unless they move quickly to fully support the census, we run 
the risk of irreparable harm.
    And frankly, I am not only worried about the problems 
presented in response rates by this controversy. I'm also 
concerned about the welfare of the hundreds of thousands of 
Americans who will be going door to door in their neighborhoods 
in the coming weeks. So today I am happy to hear things are 
going well. I sincerely hope they will continue to go well, 
despite the impact of this controversy over the long form.
    I look forward to hearing from Dr. Prewitt today on how he 
thinks this controversy will impact the census effort.
    Thank you, Mr. Chairman.
    [The prepared statement of Hon. Carolyn B. Maloney 
follows:]

[GRAPHIC] [TIFF OMITTED] T0056.004

[GRAPHIC] [TIFF OMITTED] T0056.005

[GRAPHIC] [TIFF OMITTED] T0056.006

[GRAPHIC] [TIFF OMITTED] T0056.007

[GRAPHIC] [TIFF OMITTED] T0056.008

[GRAPHIC] [TIFF OMITTED] T0056.009

[GRAPHIC] [TIFF OMITTED] T0056.010

[GRAPHIC] [TIFF OMITTED] T0056.011

[GRAPHIC] [TIFF OMITTED] T0056.012

[GRAPHIC] [TIFF OMITTED] T0056.013

[GRAPHIC] [TIFF OMITTED] T0056.014

[GRAPHIC] [TIFF OMITTED] T0056.015

[GRAPHIC] [TIFF OMITTED] T0056.016

[GRAPHIC] [TIFF OMITTED] T0056.017

[GRAPHIC] [TIFF OMITTED] T0056.018

[GRAPHIC] [TIFF OMITTED] T0056.019

[GRAPHIC] [TIFF OMITTED] T0056.020

    Mr. Miller. I am sure that you are pleased to see the 
public service announcement that Senator Lott and 
Representative Thompson put together to encourage 
Mississippians to complete their form.
    Mr. Ryan.
    Mr. Ryan. Thank you, Mr. Chairman. I wasn't planning on 
doing an opening statement, but given the controversy and 
discussion over the long form, it is prudent to make some 
suggestions.
    I am doing a PSA for the State of Wisconsin with my 
Democratic colleague from Milwaukee, Tom Barrett, urging 
everyone to fill out all of their census forms. I agree with 
you that, and as a person who believes in limited government, I 
think it is very important that you fill out the census forms.
    You heard a lot about this on talk radio, and a lot of 
letters that I am getting in my office are, ``why do they want 
to know so much about me?'' A lot of the talk radio hosts--and 
I think it is a simplistic, but interesting way of looking at 
it--say, ``if you want the government to do everything, then 
they need to know everything about you.'' That is the simple 
thing, and we are hearing that throughout the country today. We 
are hearing it more in the year 2000 than in 1990, I think, 
because there are more legitimate privacy concerns related to 
the technology that we have in this country today.
    E-commerce, the Internet, these things I think are 
symptomatic of the new technologies that are emerging in our 
economy and our society that are cause for a rise in personal 
privacy concerns. So I am not sure that this is all some kind 
of asperity against our government, but more a general concern 
about privacy rights that is rising throughout the entire 
country.
    These are basically the same questions that we had in 1990. 
It is a different country now in the year 2000, but I hope we 
can get through this and learn some lessons on the long form. 
Now that we are in the information age, hopefully we can take 
some lessons from this long form issue on a bipartisan basis 
and work forward to make sure that the next census addresses 
these privacy concerns. I think it is important that everyone 
fills out every part of the questionnaire.
    I look forward to hearing your testimony.
    In Wisconsin we had a 59 percent initial response rate, and 
we are proud of that. The reports are showing that you are on 
your way.
    Mr. Miller. You had to bring up that they beat Florida, 
didn't you?
    Mr. Ryan. Yes, sorry.
    Mr. Miller. If you would stand, Dr. Prewitt, and the three 
senior staff members with you, Mr. John Thompson, Mr. Marvin 
Raines, and Mr. Bill Barron, will also be sworn in case they 
are needed to answer questions.
    [Witnesses sworn.]
    Mr. Miller. For the record, all four answered in the 
affirmative.
    Director Prewitt, would you proceed with an opening 
statement.

  STATEMENT OF KENNETH PREWITT, DIRECTOR, U.S. BUREAU OF THE 
 CENSUS, ACCOMPANIED BY JOHN THOMPSON, MARVIN RAINES, AND BILL 
               BARRON, U.S. BUREAU OF THE CENSUS

    Mr. Prewitt. If I may preference my opening statement with 
a statement of sympathy for the unhappy evening that you spent 
Monday night.
    Mr. Miller. At least we made it into the finals.
    Mrs. Maloney. I thought you were talking about our homeless 
count night.
    Mr. Miller. The Florida Gators.
    Mr. Prewitt. Mr. Chairman and Mrs. Maloney and members of 
the committee, when I last testified, the focus was on whether 
the Census Bureau could pull off the many complex and massive 
operations--all of these operations were conducted successfully 
with no major problems that would put the census at risk.
    In your letter of invitation, you ask for the status of 
nationwide mail response rates and what those rates translate 
into for the nonresponse followup [NRFU], workload, hiring and 
other operations, and associated costs. As of this writing, the 
national mail response rate as posted on the Internet is 55 
percent. In a few hours, we will update it to 57 percent.
    It does not reflect what we expect to be an April 1 effect. 
We are not yet certain, but we are cautiously optimistic that 
we will achieve the 61 percent on which we based our budgeting 
and staffing program. April 11 is the cutoff date for 
identifying housing units that have not mailed back a 
questionnaire so we can include them in the nonresponse 
followup workload. We will continue to process mail returns 
after that date. On April 17, we will produce a late mail 
return file that we will transmit to the Local Census Offices 
so they can delete those addresses from their nonresponse 
followup assignments.
    You asked, sir, for an update on the status and a brief 
overview of the census 2000 operational time line, and 
readiness for key activities and dates that lay ahead. On many 
of these issues, the GAO will be testifying, and thus I will be 
very brief.
    We began and completed the update/leave operation as 
planned.
    Telephone questionnaire assistance centers also began on 
March 3 and will run through June 8, and outbound calling from 
the TQA sites as part of our coverage edit program will 
continue into mid-June. We have answered nearly 6 million 
calls. Just over 4 percent of those calls were unable to get 
through; almost all of those were on the first 2 days. There 
were also some early problems in validating the questionnaire 
data that was taken over the telephone. These problems have now 
been resolved. The advance letter provided an opportunity for 
those who want a language form. We have received about 2.5 
million such requests.
    In the mail out/mail back areas of the country there were 
some households that received duplicate questionnaires. This 
occurred because during all of the overlapping processes used 
to build the master address file, we wanted to minimize the 
chance that we would eliminate an address that should be 
retained. We have procedures in place to eventually remove 
these duplicate addresses from our files before the final 
census data are tabulated.
    Enumerators are visiting about half a million housing units 
in list/enumerate areas, an operation similar to that initiated 
in Alaska on January 19.
    Last week, we completed the Service-Based Enumeration. 
Census enumerators interviewed people in shelters, at soup 
kitchens, mobile food van stops and at targeted outdoor 
locations. We enumerated about 22,000 such places over the 
course of the 3 days.
    We have initiated the transient night operation, which will 
extend until April 14 for a few very large and relatively 
stable locations. We have initiated, and will continue through 
May 6, the count of about 7 million people in about 125,000 
special places during group quarters enumeration--college and 
university dormitories, hospital and prison wards, migrant farm 
camps and nursing homes. We are on schedule with regard to the 
enumeration of land-based and shipboard military personnel and 
people aboard U.S. flag-bearing merchant vessels, about 1,000 
ships and over 500 military reservations in all.
    In your letter of invitation, you asked about the status of 
data capture systems for all four sites. Data capture is 
working very well. We have scanned about 24 million forms, and 
scanning accuracy is exceeding expectations.
    We have received nearly 60,000 responses through the 
Internet.
    Questionnaire Assistance Centers opened on March 8 and will 
be open through April 14. To maximize use of staff, we have 
eliminated redundant sites and currently have 24,000 in 
operation.
    Be Counted Forms became available on March 31 at 
approximately 19,000 sites in addition to the QACs, where they 
are also available.
    Your letter also asked about any difficulties confronting 
Local Census Offices. None of the 520 LCOs is experiencing 
problems that have prevented normal operations. Some LCOs are 
reporting minor problems with their telephone systems, and 
headquarters staff are working closely with the General 
Services Administration and telecommunications service 
providers to resolve the problems. At present, all systems are 
up and running.
    Nonresponse followup [NRFU], is scheduled to begin April 
27. Enumerator training begins April 24, and NFRU will continue 
for 10 weeks until the first week of July. Extending NRFU 
beyond that date would not only increase census costs, it could 
lead to a reduction in data quality. Experience teaches us that 
the longer we are in the field, and the farther we get from 
census day, the more the quality of respondents' answers 
deteriorates. We will stay in the field until we have exhausted 
all of our established procedures.
    You asked about the status of the hiring process for NRFU. 
While we have met our national goal of having 2.4 million 
qualified applicants well in advance of our April 19 target 
date, we are continuing to accept applications and to actively 
recruit in local areas where we have not yet met our recruiting 
goals.
    I would now like to describe in some detail the 
enumerator's job and our procedures for assuring the quality 
and completeness of their work. Each NRFU enumerator is 
assigned a specific area in which to work, called an assignment 
area, and is given a binder of addresses in that area that 
includes all those addresses for which we have not received a 
completed questionnaire, and in rural areas enumerators also 
receive maps that have the housing units' locations spotted on 
them.
    If the current household lived at the address on census 
day, the enumerator interviews a household member at least 15 
years of age and completes the assigned questionnaire. If the 
unit was occupied by a different household on census day, the 
enumerator completes a questionnaire for the occupants who 
lived there on census day by interviewing a knowledgeable 
person, such as a neighbor. If the current occupants were not 
enumerated elsewhere, the enumerator will also complete a 
census questionnaire for them at their census day address. If 
the housing unit was vacant on census day, the enumerator 
completes appropriate housing questions on the questionnaire by 
interviewing a knowledgeable person, such as an apartment house 
manager.
    The enumerator must make up to six attempts to complete a 
questionnaire. If no one is home at a housing unit, the 
enumerator obtains as much information as possible about how to 
contact the occupants. The enumerator leaves a notice at the 
address that they have been visited and provides a telephone 
number so the occupant can call back. He will make up to two 
additional personal visits, three in all, and three telephone 
attempts at contacting the household before obtaining as much 
information as possible to complete the questionnaire from a 
knowledgeable source.
    Enumerators are instructed to make their callbacks on 
different days and at different times of the day. They must 
obtain at least the status, occupied or vacant, and the number 
of people living in the unit. If an enumerator submits a 
questionnaire which contains that minimal level of data, the 
crew leader must check the enumerator's record of callbacks for 
the housing unit to determine that procedures were properly 
followed. The crew leader also holds these cases for possible 
further followup to obtain more complete data.
    In order to prevent falsification of the data by 
enumerators, a percentage of each enumerator's work is verified 
for accuracy by staff. An enumerator who is discovered 
falsifying data is dismissed immediately, and all the work must 
be redone by another enumerator.
    Daily production levels begin to decrease during the end of 
NRFU. Sometime enumerators complete the easiest cases first, 
finish the work closest to their homes first, or believe that 
the quicker that they finish, the sooner they would be out of 
work. In order to bring the NRFU to closure within schedule, we 
implement a procedure known as ``final attempt.'' Within the 
area covered by a crew leader, approximately 2,200 cases, when 
that area has completed 95 percent of its workload, the crew 
leader consolidates the remaining work and gives it to the most 
productive and dependable enumerators. They make one final 
visit to each outstanding address and do some of the housing 
units for which only minimal data was collected to complete as 
much of the questionnaire as possible. This procedure takes 
advantage of our best enumerators and will improve both the 
count and the data quality.
    Final attempt must resolve all outstanding cases. NRFU is 
not over until every procedure has been completed, and this, of 
course, includes the check-in of every census form.
    Let me then turn quickly to the long form issue.
    Mr. Chairman, I pledged to you and this subcommittee 
several meetings ago that I would bring to your attention any 
development which could put the census at risk. Nothing in our 
current operations poses such a risk, but the widespread attack 
on the long form could have serious consequences. Indeed, I 
alerted you to this in our phone conversation early last week. 
First a few background comments.
    Concern with overburdening respondents with too many 
questions led the Census Bureau to introduce a long form on a 
sample basis in the 1940 census. We have used this approach in 
each decennial census since. The selection of a sample based on 
established scientific methods means that not everyone is asked 
every question. The majority receive only the short form.
    The census 2000 long form is the shortest in history. The 
law requires that 3 years prior to census day, the Census 
Bureau report to Congress the subjects proposed for inclusion 
in the census. The Census Bureau reported this information to 
Congress in a letter accompanying materials dated March 28, 
1997. The law also requires that we report to Congress the 
specific questions we intend to ask 2 years prior. We did that 
March 30, 1998. The materials that we submitted to Congress 
described each question we included on the long form and, more 
importantly, provided detailed legal citations that indicate 
each item is mandated or required by congressional legislation 
or Federal judicial decisions in the book that the ranking 
member and indeed you referenced as well.
    Accurate census data provide the underpinnings for other 
Federal surveys and data collections. The decennial census 
forms a sampling base for other national surveys and is used to 
compute rates of various indicators. Therefore, it is directly 
linked to the statistical system's ability to provide current 
unemployment data, to provide data for making cost-of-living 
adjustments, to calculate numerous vital statistics and rates 
for health services, to calculate crime and victimization rates 
and the like.
    I now bring the subcommittee up to date regarding our 
concerns about the fate of long form data in the current census 
environment. Some of the information I now have available is so 
recent that I could not include it in the written testimony 
submitted earlier this week.
    The current differential response rate between the short 
and long form household is approximately double the 1990 rate. 
This differential may close, and we are doing everything we can 
to assure the American people that long form data are important 
and confidential. Every 5 percent differential in the response 
rate between the two forms translates into a 1 percent 
reduction in the overall response rate. In other words, if a 
differential today were what it was in 1990, the overall 
national response rate would be a percentage point higher.
    If the lower than expected response to the long form 
persists, there will be operational and budgetary implications. 
It takes more time to enumerate a long form. A lower than 
expected response rate will, consequently, place an 
unanticipated burden on the nonresponse followup phase of the 
census. Moreover, given the public atmosphere that has 
trivialized and discredited the long form, we have to be 
concerned about the morale of the field staff who will now be 
trying to get information that many public voices, including a 
few Members of Congress, are saying should be voluntary. We 
have to be prepared for higher than expected turnover, 
especially in rural areas with the higher than average number 
of long forms.
    Given the public commentary, there is also the possibility 
that we will have a higher than expected item nonresponse on 
the long form. This could have serious consequences for a 
decade. The Census Bureau has high quality standards. It would 
not release data that it believed were insufficiently reliable 
to perform the functions expected of them. This has never 
happened with census data, but it has with certain survey 
information. If the two issues just mentioned--high nonresponse 
to the long form and high noncompliance with particular items 
on the forms returned--combine to push data below our quality 
threshold, the Census Bureau would be placed in a very 
difficult position of deciding what to release.
    Mr. Chairman, I know you are concerned about whether the 
ACE will provide the quality of data required to adjust for the 
undercount. At a public session organized by the National 
Academy of Sciences, I said if the ACE effort did not meet 
Census Bureau quality standards, it would not be used. This 
holds for all Census Bureau efforts. If, for instance, the 
income data were to fall below our quality threshold and we 
could not release it, more than two dozen statutory uses 
ranging from the Energy Policy Act of 1992 to the Business and 
Industry Guarantee Loan Program of 1980 to title I funds and 
Head Start programs would be affected. So also would be the 
calculation of the Consumer Price Index and the unemployment 
rate for the next decade.
    You, Mr. Chairman, and the ranking member and Mr. Ryan and 
Mr. Davis have made strong statements about the importance of 
the long form data, but now I urge you to ask the entire U.S. 
Congress to step forward and explain to the American people why 
the Congress has required, authorized and paid for the 
collection of these long form data. There were no viable 
alternatives to having a long form for census 2000. No other 
data source could provide all the information that a Nation 
needs in a cost-effective manner. In the long term, we hope 
that the American community survey will replace the long form, 
and indeed by 2010. The ACS scheduled for nationwide 
implementation in 2003 is one of the most important 
improvements in Federal statistics, and it is the cornerstone 
of our efforts to keep pace for timely and relevant data.
    Thank you, Mr. Chairman.
    [The prepared statement of Mr. Prewitt follows:]

    [GRAPHIC] [TIFF OMITTED] T0056.021
    
    [GRAPHIC] [TIFF OMITTED] T0056.022
    
    [GRAPHIC] [TIFF OMITTED] T0056.023
    
    [GRAPHIC] [TIFF OMITTED] T0056.024
    
    [GRAPHIC] [TIFF OMITTED] T0056.025
    
    [GRAPHIC] [TIFF OMITTED] T0056.026
    
    [GRAPHIC] [TIFF OMITTED] T0056.027
    
    [GRAPHIC] [TIFF OMITTED] T0056.028
    
    [GRAPHIC] [TIFF OMITTED] T0056.029
    
    [GRAPHIC] [TIFF OMITTED] T0056.030
    
    Mr. Miller. We have two votes coming up, and so I think we 
should be back in about 15 minutes. We stand in recess. I ask 
my colleagues to come back as soon as we can, and we will 
proceed.
    [Recess.]
    Mr. Miller. We will reconvene the subcommittee. Let me 
start off with some questions on the long form.
    What is the difference in response rates in 1990 between 
the long form and short form, and also in the dress rehearsal?
    Mr. Prewitt. The long form/short form differential in 1990 
at the end of the census was 4.5 percent, but at the end of 
mail out/mail back, it was about 6 percent. The reason that 
converged slightly, when we went out in the field, we were able 
to convert a higher percentage of the long from nonrespondents 
than the short form nonrespondents, so we closed the gap in 
1990.
    Your numbers that I just saw in your testimony on the dress 
rehearsal ranged from 10 to 15 percent.
    Mr. Miller. It was Sacramento, and I don't think----
    Mr. Prewitt. Sacramento and South Carolina. Sacramento was 
14.7--12 percent, and South--that's South Carolina.
    I'm sorry, the reason that it is complicated, we calculated 
both the mail out/mail back and update/leave area. So the 
update/leave area was 13 percent. The mail out/mail back area 
was 11 percent for South Carolina.
    The differential in Sacramento for mail out/mail back was 
15, and in Menomenee was 8. That was all update/leave. Those 
are the numbers.
    Mr. Miller. So the dress rehearsal gave us an indication of 
a problem which we just found out about a year ago, and at that 
time it was too late to respond to it as much. What steps did 
the Bureau take?
    Mr. Prewitt. I would say that there are certain things the 
dress rehearsal gives you a clue on. As you know, the overall 
turnout response rate in dress rehearsal was low. It doesn't 
predict everything. It is an opportunity for us to test 
operations. We don't expect the response patterns in a dress 
rehearsal to look like the overall response rates. We would not 
ourselves have concluded that that differential was very 
predictive.
    We thought the strongest predictor of large-scale patterns 
is the 1990 pattern. Indeed, one of the most interesting things 
is that the overall response rate in 1990 compared to 1980 
tracks almost perfectly across the 50 States. It is just that 
everybody dropped 10 percent. It is not that some States 
dropped 20 and some States didn't drop at all; all dropped 
approximately 10 percent across the country. That is the 
strongest predictor. We based much of our operational 
predictions on the 1990 response rates for 2000.
    There are so many things going on in a dress rehearsal. 
One, they are not typical places of the entire country.
    Mr. Miller. There was a large differential. You don't think 
that was significant in both Sacramento or----
    Mr. Prewitt. No, we didn't conclude from that we were going 
to get this kind of differential in 2000, but neither did the 
subcommittee or GAO. Nobody said, oh, my goodness, at that 
stage.
    Mr. Miller. When you scan in the envelopes the bar code 
tells you whether it is a long or short form. You don't know 
whether the person completed just the first six questions?
    Mr. Prewitt. That's correct. We will do serious work on 
item nonresponse, but we won't have serious data until during 
the winter of 2001.
    Mr. Miller. I was talking to a Member of Congress, and he 
had the long form. He was still completing it. I got the long 
form, and there are some questions my wife had to fill out 
because she knew more details. The short form is--obviously 
anybody can go through it in a couple of minutes and complete 
it. There could be a delay a little bit, so we will have to see 
what it is.
    Mr. Prewitt. We very much hope that there is a delay, and 
we hope that people are sitting with the long form waiting and 
that this converges.
    If you do the arithmetic, there aren't that many forms left 
out there that we expect to get back in the mail. At a certain 
point you begin to get a real tailing off. We are hoping that 
this weekend--we are doing a lot of heavy advertising. It is 
certainly possible, as you suggest, Mr. Miller, that more long 
forms are sitting on those kitchen counters, and we will get a 
disproportionate number of long forms at the tail end. And we 
will be happy if that turns out to be the case, but we will 
know that roughly a week from today.
    Mrs. Maloney. Thank you, Mr. Chairman.
    Dr. Prewitt, I think the controversy of the long form that 
has surfaced has been quite harmful to your efforts. What do 
you think is the impact on the response rate because of these 
comments by elected officials?
    Mr. Prewitt. I honestly believe that it is very difficult 
for large parts of the American public to draw the kind of fine 
distinctions that are sometimes suggested in public commentary.
    I appreciate that all responsible leaders are saying it is 
important to be counted; therefore, send your form in even if 
you don't fill it all out. But how that translates in the 
public consciousness, especially since we are now dealing--we 
have all got to remember, we are now dealing with the tail end 
of the mail-back response period. That is, the most alert and 
responsible and committed members of society have probably sent 
forms back in. So we are now dealing with people who are less 
motivated or less attuned and paying less attention. What they 
may hear vibrating in the atmosphere is, ``oh, well, the 
information is not that important after all.'' That is what has 
us worried.
    Mrs. Maloney. What are you doing to counter this 
unfortunate attitude? Do you have any plans to specifically 
respond to the unfortunate comments of Senator Lott and 
Governor Bush?
    Mr. Prewitt. I have done everything that I can in the media 
to repeat that the long form questions are all there because 
the U.S. Congress wants them; that they all perform these 
important functions, as you have all testified and said in your 
own PSAs; and that all we can do is simply repeat that.
    We are doing a lot of targeted advertising, video news 
feeds. I do about 10 or 15 a day where we think that we might 
be able to get a bit of visibility on this. We are accelerating 
our targeted radio advertising right now, but--it is very late 
in the game to try to use an advertising campaign to counter 
the mind-set or the public impression that has been generated 
by--and I think as the chairman says quite correctly--a quite 
extensive attention to this issue among talk shows and other 
public commentators. When I say that, I certainly don't mean at 
all to exclude any of the larger--the newspaper editors and so 
forth are all part of that commentary.
    All we can do at this stage is push hard in the last 3 or 4 
or 5 days.
    If I can say one other word, I think it is going to be 
extremely important when the mail-out/mail-back period is 
finished, which is, after all, less than a week, to regroup on 
this and try to get a message out, because the nonresponse 
followup period, we are going to have a lot of temporary 
employees, they are Americans trying to count America, and they 
are going to be out in the field knocking on doors, and it is 
very important to have an atmosphere at the time that this 
census matters, that this is serious business, and that this is 
not trivial or incidental or voluntary.
    So I am very much hopeful that we will be able to, with 
your help, enlist the U.S. Congress on that behalf, and other 
members of the U.S. Government, to say--we may have another 40 
to 45 million households. So we have another shot at trying to 
make a major message, but we will not be able to do that 
through an advertising campaign. We will have to do that with 
the kind of PSAs you just saw, and I hope they will stress the 
importance of these data and to cooperate with the enumerators.
    Mrs. Maloney. I must say that I have collected well over 30 
editorials across the country really calling upon everyone to 
fill out their forms, the long form, and not to listen to any 
elected official who may be advocating otherwise or referring 
to the census long form as optional.
    It occurs to me that the problem may surface after the 
mail-back, but in the nonresponse followup. It may be more of a 
problem there. At what point do you send an enumerator out, 
once you have the long form? Do all of the questions have to be 
answered? What is the decision if they do just selectively 
answer; do you send out an enumerator? What is the procedure in 
that case?
    Mr. Prewitt. No, if we get a long form in that has any 
information whatsoever that allows us to consider it a 
legitimate response, then we cannot send an enumerator out to 
try to get the additional information. That is why I say item 
nonresponse is a very serious issue, but we don't have a good 
measure. It could be three questions left blank, or it could be 
52 questions that were left blank.
    We certainly have to have some information. For example, we 
cannot take a form that says there are 99 people living here 
and then nothing else. We can't accept that form on behalf of 
the U.S. Government. We would have to somehow find out if there 
were really people living there.
    So there are certain thresholds below which we cannot 
accept the form, and you wouldn't want us to. It would be an 
alert to us that perhaps this is a fraudulent count. So we have 
to get enough information to know that somebody actually lives 
there, that this is a residence, it is an inhabited residence, 
and enough information about an individual to be able to say 
this is a person or else we can't put them in the count.
    We certainly don't have the resources to go out and now 
convert a lot of empty responses on the long form into full 
data. That is not part of the census operational plan.
    Mr. Miller. Mr. Ryan.
    Mr. Ryan. Dr. Prewitt, you said that the nonresponse 
followup for the long form is twice what it was for 1990 at 
this time?
    Mr. Prewitt. At this time.
    Mr. Ryan. Why do you think that is, aside from comments 
here and there?
    Mr. Prewitt. Look, I am trying to actually get some 
information on this, and I can speculate the way that you can 
speculate. I think you are right, Congressman Ryan, that this 
country has a heightened sense of privacy concerns, and that 
spills over into the government.
    I can tell you based upon some survey data that the 
proportion of the American public who was telling us that the 
census data are invasive jumped by 7 percent from--from week 2 
of the census to week 3, and in between that period of time, 
that is when this campaign started. So I can only infer from 
that that it is having some effect. Does that translate into 
nonresponse? I can't tell you that yet.
    Mr. Ryan. I think it was a Houston judge that filed an 
injunction against the imposition of a fine for those who may 
not fill out all of their long form. What is your reaction to 
that? In 1990, did the Census Bureau impose a $100 fine on 
people who didn't fill every bit of their long form 
questionnaire? What is your take on the injunction?
    Mr. Prewitt. The last case that was enforced on 
noncompliance with the census was in 1960. Mr. Rickenbacker. 
The fine was imposed. It was upheld by the courts.
    The Census Bureau itself is not an enforcement agency and 
would never enforce any of these. We are a statistical agency. 
But it has not been our recommendation that enforcement action 
take place. My own concern on that would be that that would 
create more noise, more fuel, and I would worry that it would 
have a damaging effect on the census.
    By the way, the $100 which has been mentioned in the press, 
and indeed we have mentioned it ourselves, I want to correct 
the record, it turns out to be up to $5,000. The standard 
Criminal Act of 1984 trumps all other acts. It is title 18, I 
believe, and unless you explicitly exclude some Federal 
infraction from the law of title 18, the fine is actually up to 
$5,000.
    Mr. Ryan. I thought it was $5,000 if a government employee 
misuses the census data or accesses it improperly.
    Mr. Prewitt. That is a separate issue.
    Mr. Ryan. So the fine is actually $5,000?
    Mr. Prewitt. Up to.
    Mr. Ryan. Up to $5,000.
    Mr. Prewitt. This is the uniform criminal statute passed in 
1984 that basically, as I understand it, says that any 
infraction of a Federal law can be--can elicit a fine up to 
$5,000. So the particular injunction against the $100 is 
targeted on title 13 rather than title 18.
    Mr. Ryan. So the injunction really is meaningless. And an 
infraction subject to the $5,000 fine could be failure to fill 
out one or two of the questions on the long form?
    Mr. Prewitt. Yes.
    Mr. Ryan. I don't want to create some hysteria on talk 
radio on this. Hopefully C-SPAN will play that. The Census 
Bureau--these fines have not been imposed in the past?
    Mr. Prewitt. Since 1960.
    Mr. Ryan. They were not imposed in 1970, 1980 and 1990?
    Mr. Prewitt. I think maybe there was one case in 1970. I am 
almost certain in 1970 there was a case that was overturned. It 
was overturned on the grounds that it was selective 
enforcement. ``Why did you choose that person instead of that 
person when millions performed the infraction.'' The only one 
that was upheld was 1960.
    Mr. Ryan. So the last one was thrown out?
    Mr. Prewitt. I believe so. But the Census Bureau is not 
interested in pursuing enforcement action.
    Mr. Ryan. I understand that it is not in your best interest 
to broadcast that, because then you encourage people not to 
fill these out. Boy, that is an intriguing number.
    As your enumerators are going out--and I know that you 
addressed this with Mrs. Maloney, but as they are going out and 
following up for the long form, as they ask questions on the 
followup for the long form, is there a threshold in the 
questioning that is acceptable and then not acceptable? Meaning 
if you find that people are not going to answer a question A, B 
or C, but they will answer all other questions, is there a 
threshold in the long form that makes it acceptable census data 
or unacceptable census data? Has that threshold been 
established?
    Mr. Prewitt. There is certainly a minimal threshold. We 
have to be able to be certain that the number of people we are 
counting in this household on this block actually live in that 
household. That is the threshold.
    Mr. Ryan. So essentially the short form questions and----
    Mr. Prewitt. Yes. If we got even a partial short form 
answer on the long form, the person would still be counted. So 
we would have huge item nonresponse, but we would not lose the 
count. And we will do everything we can to get that count 
correct.
    Mr. Ryan. Thank you.
    Mr. Miller. Mr. Davis.
    Mr. Davis of Illinois. Thank you very much, Mr. Chairman.
    Dr. Prewitt, I know that there has been and continues to be 
a tremendous amount of discussion about the long form and the 
response.
    I do believe that people begin to feel that it was more 
invasive as they heard other people suggest that it was 
invasive. I mean, the power of suggestion is amazing still in 
this country. And I don't think people were concerned as much 
about whether or not it was invasive until they began to hear 
public figures suggest that maybe it was, or they saw some 
columnist suggest that maybe it was. They pick it up and say, 
``yes, I guess it really is,'' when they look at it.
    Let me just ask you, let's say that I am one of these 
individuals who want to participate in the count, and I don't 
have any real difficulty giving the basic information, but I, 
too, have been convinced, if I was that person--and I received 
the long form, and I was not convinced--I did half, and my wife 
did the other half, and then there might have been a question 
or two and we threw up a coin to decide which one of us would 
answer that one, and it was done. A lot of fun.
    But let's say that I am not convinced that the information 
is necessary, and that I can participate without providing this 
information. Is there something that one might be able to 
suggest or convey to the average citizen that it is important 
to do the long form if that is what they got?
    Mr. Prewitt. Well, Congressman Davis, I think that the 
message is roughly the message that we have been trying to 
promote now for 6 months, which is an awful lot of government 
programs, provide benefits to your community, if you see those 
ads about schools, and you see those ads about transportation, 
or you see those ads about day care centers, if you make any 
kind of connection, you connect that to long form data, because 
all of the social programs use the kind of data about age, 
about veteran status, about poverty, about traffic congestion, 
about water pollution, based on long form data to provide those 
services.
    I would hope that when you are sitting there at the table 
and saying, I know this is something that I don't want to do, 
but I have just heard that all of these benefits will come to 
my community, you will make that sort of logical step.
    But at this stage what we will have to do--because if we do 
have a higher than expected nonresponse to the long form, we 
will now have to try to get the enumerators--and this is not 
easy. You are trying to train half a million temporary workers 
to enumerate people who are angry at you, indifferent, hostile 
toward you and get that full information. We will have to rely 
on that army of people. We will have to get them to understand 
the importance of this. This will not recapture the data that 
has already come in, but is incompletely filled out. There is 
no way to recapture that data at this stage.
    Mr. Davis of Illinois. Are you saying that this is 
information that can be used for planning purposes to help make 
specific determinations about what is needed in certain 
communities or what might be needed overall for the country as 
a whole?
    Mr. Prewitt. Well, yes, sir.
    To put this as strongly as I can, I think the commentators 
thus far are overlooking the fact that the Consumer Price 
Index, the unemployment rates are tracked with data that in 
turn are dependent upon quality information that you get from 
the decennial, and we are putting at risk the way that we 
conduct our basic economic statistics in this country. This is 
very serious stuff.
    Mr. Davis of Illinois. Would you also say that there is no 
better way or no other time at which we could expect to get 
this information in such a massive way?
    Mr. Prewitt. There certainly is no way in the year 2000. 
There is no agency other than the Census Bureau that can 
collect this kind of information. We cannot suddenly decide 
let's find somebody else to collect long form data for the 
country.
    The best we could offer the country, and it is not trivial, 
is that if we were to--the chairman in his opening remarks said 
he does not expect to be doing the long form data ever again, 
holding out the possibility that we will be able to launch the 
American community survey. We are currently scheduled to launch 
that in 2003. We could actually accelerate that by a year. We 
could start the American community survey a year earlier if the 
Congress instructed us. If they told us to start planning to be 
in the field by 2002 with the American community survey, I 
believe we could do that.
    Mr. Davis of Illinois. I know that Illinois' initial 
response rate is 56 percent, but I am told that my city is 
significantly lower than that. Would you have any suggestions 
at this late date for those places that are coming in below the 
national norm?
    Mr. Prewitt. The most important figure to watch right now 
is how far below your own performance in 1990 you are. In 
Chicago you are about 13 percent below your 1990 performance. 
That is not that far off from the national number. The national 
number is about 10 percent. Even though you are well below the 
national average, the most important thing is to measure 
yourself against 1990.
    So the most important message to get to the people of 
Chicago is let's accomplish what we accomplished in 1990. Worry 
about what we were in 1990 and how we can get there. It is not 
too late to send the form in. We are now doing video news feeds 
to Chicago saying it is not too late, it is not too late. Mail 
it back now if you still have it. I think the more we can get 
that message out over the next 2 or 3 days, the better off the 
census will be.
    Mr. Davis of Illinois. Thank you. I am so pleased I have a 
bunch of volunteers who are also going out this weekend simply 
knocking on doors and asking people to send their forms in.
    Mr. Prewitt. Good, good, good.
    Mr. Miller. Director Prewitt, I have been urging people to 
complete all of the questions because we recognize how critical 
it is for our area. Sarasota is undercounted, and it hurts; 
Chicago or New York City. So it really is a personal thing that 
we need to do.
    When Mr. Ryan was asking the question about why people are 
not responding, you referred to some poll that said it is 
really because of some comments of Andy Rooney or politicians 
or all of the talk show people. There are legitimate concerns 
about privacy that are probably different today than 10 years 
ago, whether it is medical privacy--financial privacy is always 
a subject that we are concerned with, and we have legitimate 
concerns.
    I mentioned in my opening statement a problem with the 
abuse of driver's license lists in Florida. They were selling 
photographs even in Florida, so people are more suspect of 
government. So it is not just these comments. There are 
differences in society.
    Mr. Prewitt. Yes, sir. Yes, sir. I don't want to back away 
from that.
    What I have said publicly and repeat strongly today, I 
think this country is on a collision course between its 
insatiable desire for information and its heightened concern 
for privacy, and the Census Bureau is caught between those two 
needs. As I said yesterday publicly to the press, we are 
creating a knowledge economy, and the infrastructure for a 
knowledge economy is information. And the decennial base helps 
create a higher quality information infrastructure for this 
society, and the society on the one hand wants that, and on the 
other hand we have these deep concerns about privacy.
    All I was suggesting by the poll data, and I don't want to 
put too much emphasis on poll data, but in a week--it wasn't 
that it wasn't already there, it was, but it jumped in 1 week 
that the census data are invasive, and it happened to be the 
week that this became a public discussion. That is a fact. I 
don't want to overinterpret it.
    Mr. Miller. I think if we polled it today compared to 10 
years ago, it would be higher. We are in an information 
technology era, and it raises these concerns. After we get 
through these critical phases, we are going to discuss how to 
handle the 2010 census.
    If someone refuses to answer the income question, and you 
get asked this question, what do you tell someone? You tell 
them basically--what do you tell someone who says, I am not 
going to put down how much my electric bill is?
    Mr. Prewitt. One on one with a respondent, I would say, 
look, give us an estimate, create a range, give us the best 
information that you are prepared to give us. Here are the kind 
of ways that this information is used. As I just said, over two 
dozen pieces of important Federal legislation use some--the 
income data one way or the other. So the array of programs that 
use these data is enormous. But it is also used to drive the 
sample frame and the statistical controls for the CPI and 
unemployment data. All of our pension systems are indexed to 
the CPI. The Social Security is indexed to the CPI. The stakes 
are very high. That is what I would try to explain.
    If they persisted in refusing, I would prefer to get their 
information, whatever I could get from them. The most important 
thing--and I don't underestimate this--the most important thing 
is a good count. Our constitutional obligation is to count the 
population for purposes of apportionment and redistricting. We 
take that as our foremost priority task; and the other benefits 
that come from the long form are simply not as high a priority. 
So we will do everything we can to count everyone and make sure 
that we don't count anyone twice and that we have no fraudulent 
responses. That is our first task.
    Mr. Miller. You are not an enforcement agency, as you said 
to Mr. Ryan? You are not an enforcement agency?
    Mr. Prewitt. We are not going to tell our enumerators to 
wave fines in front of these people. We did put on the envelope 
that it is required by law. We wanted to make sure that this 
does not look like junk mail. We were worried that people might 
try to duplicate the census mailing, as indeed we had one 
instance of, and indeed that mailing must have worked because 
we got some checks made out to that organization.
    Mr. Miller. What did you do?
    Mr. Prewitt. We sent them back to the respondents. We 
didn't want to be in the banking business of handling money for 
Mr. Glavin, as you might appreciate. We actually did get some 
responses to that mailing.
    But by putting the mandatory nature on the envelope, we 
were certain that nobody could duplicate the envelope and try 
to piggyback on the census environment. The other reason is 
that we have some research that suggests that slightly 
increases the response. We wanted to use everything that we 
could to get the response rate up.
    Mr. Miller. Did you get anyone that sent you a check and 
said, I refuse?
    Mr. Prewitt. Oh, yes. We have certified letters that come 
in with $100 saying, I am going to pay my fine.
    Mr. Miller. But the check has an address?
    Mr. Prewitt. Listen, the number of things we get, you would 
be surprised. The other day we opened up a form, there were 
seven $100 bills in it, and obviously somebody made a mistake. 
They had stuff on their desk that got put in. We found that 
person in less than 24 hours and returned the money to them. I 
was very proud of our organization.
    Mr. Miller. Mrs. Maloney.
    Mrs. Maloney. That is a good story to tell from your 
agency.
    At our last hearing we had quite an extensive discussion on 
access by various agencies to the Census Bureau. I understand 
that many of these issues have been worked out, but that there 
are ongoing conversations with the Monitoring Board.
    I have also heard that there has been some confrontation 
between oversight personnel and Census Bureau personnel, and I 
understand there were some threatening comments.
    Could you explain what happened and comment further on 
access, and in particular this particular incident?
    Mr. Prewitt. No, I am pleased to report that we have been 
making some headway, and with the chairman's permission and 
your permission, the Deputy Director has taken a major 
leadership responsibility in working out the access questions, 
so if I can ask him to respond to where we are on access 
issues.
    Mrs. Maloney. Go ahead.
    Mr. Miller. Mr. Barron.
    Mr. Barron. Good afternoon.
    Mrs. Maloney. Good afternoon.
    Mr. Barron. Yes, we have spent a lot of time on access 
issues. I think the major objective was to make sure that we 
were providing the access that all of the various oversight 
entities felt that they needed in order to do their job. Right 
now I think we are at 140 visits, either conducted or scheduled 
now through the end of April. I fully expect that number is 
going to grow some more. To my knowledge, we are working well 
with all those who wish to look at our activities. If there are 
any complaints, I hope that people will get in touch with me 
right away.
    With respect to the issue of threatening comments, I think 
we did have reports of one incident in one LCO. I have 
discussed that with the Monitoring Board staff. I think they 
agreed with me that this was a situation that needed to be 
addressed, and, in fact have now issued some guidelines on 
conduct which emphasize that in the course of doing these 
visits, Federal employees and particularly LCO staff need to be 
treated with courtesy and respect. I think that is mentioned 
several times in those guidelines, and I would like to thank 
the congressional side of the Monitoring Board for preparing 
that document and putting this issue to rest.
    Just in conclusion, I think given the tone of some of the 
comments made at the last hearing, I think this was the reason 
the Census Bureau had our guidelines in the first place. We 
have a temporary staff working for us for just a short period 
of time. They are a wonderful group of people, and we give them 
a lot of work to do, and we were just trying to manage the 
process by which people contact them. And over the last month I 
think we have made a lot of progress, and I am hoping others 
agree and we can go about doing the work that we need to do.
    Mrs. Maloney. Dr. Prewitt, what is your response to the 
chairman's comment that he would like to do away with the long 
form in 2010?
    Mr. Prewitt. Well, I did obviously note that response or 
that comment. I agree with the chairman. I think, as the 
chairman knows, the Census Bureau has for several years been 
working toward establishing the American community survey. 
Congress has funded this early preparatory work. We are in the 
field right now to see if the questions bridge between the 
American community survey format and the long form format in 
the decennial.
    We are coming before the Appropriations Committee tomorrow. 
We will be recommending in our fiscal year 2001 budget the 
continuation of that work. I do not see any alternative to the 
long form other than the American community survey. I think 
some of the ideas that have been mentioned in public that we 
ought to simply assign this task to each of the agencies to do 
their own individual surveys would not be a very efficient way 
to conduct the government's business.
    So I do think that the American community survey remains 
the most innovative and important way to get the kind of data 
that the country needs, not just the Federal Government, but 
the country needs in a timely fashion and to do it in a 
somewhat different environment.
    The questions, I should say to the committee, are no less 
intrusive. They are still the same questions unless the U.S. 
Congress decides we should not be asking these questions, which 
is fair enough, we won't ask them. But we believe in a sample 
format in which you are only talking to a quarter million 
people per month, that you are rolling that through the full 
year and the next year and the next year, that you have the 
opportunity to do more education about the importance of these 
questions with the local leaders.
    I think when--the important thing about the American 
community survey is that it is conceptualized to be deeply 
rooted in the local communities, and when the local leaders 
understand these are important data for us, then we hope that 
they will be out front in making the case, and that will create 
a public education environment, and we will get high levels of 
cooperation.
    Mrs. Maloney. Although I was not a Member of Congress in 
1990, I was a member of the city council in New York and was 
very involved in the census and involving partnerships with the 
community and working with other Congress Members to get the 
response rate up. I don't recall any type of objection or 
conflict at all over the long form in 1990, and the form that 
we have before us now is essentially the same, only four 
questions less.
    You mentioned there was a disparity between the short form 
and the long form after the second week; is that correct?
    Mr. Prewitt. No. What I was talking about was some survey 
data.
    Mrs. Maloney. About the response rate coming back?
    Mr. Prewitt. Right.
    Mrs. Maloney. After the controversy, the response rate fell 
for the long form?
    Mr. Prewitt. No. Actually we have not tracked this day by 
day. I don't know as we would put much confidence even if that 
were the case, because as the chairman said, we expect people 
to hold the long form longer and to be delayed in returning it. 
So what we are focused upon is the end point. If we don't close 
what is now roughly a 12 percent gap in the long form and the 
short form response rate, then, as I say, operationally we have 
more work to do, and we also have the problem with data quality 
if we don't get those data.
    So the most important indicator, I think, of whether the 
campaign has had an effect will be on item nonresponse. That 
is, if we have millions of long forms that have come in, but 
there is not much on them, and if there is a significant drop-
off from 1990, then we would be able to infer that obviously 
the conversation, as Mr. Davis just said, the kind of 
suggestive nature of invasiveness will have had an effect, and 
the country will pay a price for a decade unless we get the 
American community survey in quite quickly and fill in the 
gaps.
    It is serious stuff, and I am concerned that people don't 
understand what is at stake when you are talking about the CPI 
and Social Security payments, to say nothing of title I and 
Head Start and Clean Air and all of the other programs, the 
dozens and dozens and dozens of programs. But as I have said 
publicly, I think that the capacity of Mr. Greenspan to report 
to this Congress on the state of the economy becomes an issue 
if we have very flawed long form data.
    Mrs. Maloney. My time is up.
    Mr. Miller. Is there an organized campaign against the long 
form? A lot of talk show people are going after it. There is 
not an organized effort to do it, is there?
    Mr. Prewitt. I would say that I have certainly heard the 
leaders of the Libertarian Party, that is an organization, and 
I can only tell you from my e-mail traffic that when you start 
getting the same e-mail time and time again, it suggests that 
it is not just random, and when you hear the same sort of 
things in the talk shows. It is certainly an environment in 
which it is easier to create a buzz in the public discourse 
about something because of the Internet chat rooms. We have 
people who track the chat rooms, and there is a lot of it 
there. We have Internet sites, all of those things.
    Mr. Miller. Even Andy Rooney, who is not a conservative, 
came out saying--this is more local with me. In Sarasota, I 
think it was 58 percent as of yesterday, and I was rather 
pleased that my main county is--but the Complete Count 
Committee has received hundreds of calls from people who have 
not received a questionnaire. These are not communities with 
new housing units. There have been reports in the Washington 
Post that local areas have not received their forms. What can 
these people do to make sure that they get counted?
    Mr. Prewitt. Obviously every time we get a report that some 
area of the country has not received forms, we go to work on 
that. If we get a report that these people got their advance 
letter and their reminder card and did not get a form, for some 
reason the postal service did not mail the form. So we hope 
that those forms are sitting someplace in a post office and 
they are still in the mail stream and they will get there.
    But when you have a situation where no one got any piece of 
mail, then that suggests that there was a mail address problem. 
And if that is in new construction, we have finished our new 
construction work. We are adding about 375,000 addresses 
through the new construction process, and they will be 
enumerated in the nonresponse followup period.
    We have to figure out first what is the nature of the 
problem. You can still order a form up to April 11 by using 
that number. We widely publicized that number. We sometimes 
deliver them ourselves if we have reason to believe that it was 
a breakdown in our system. We are not finding many instances 
where it is a breakdown. Sometimes it is a slippage between the 
Post Office box problem. We cannot deliver to a Post Office box 
because that is not a geocoded address, and so some of the 
instances that we are picking up in the press and other ways 
are examples of those. But we do not ignore those. Every one of 
those we immediately, through our Local Census Offices, go to 
work in that neighborhood and sort out the nature of the 
problem and correct it.
    Mr. Miller. There is an area of Laurel in Sarasota County 
that said they were not counted. We are sending letters to make 
sure that people are aware that they will be followed up on, so 
there is a concern.
    In Florida we have a lot of seasonal residents. Longboat 
Key has a separate set of numbers, for example, but they have 
large mobile home parks for 6 months of the year. First of all, 
residents feel they should be counted half in each State. If 
one lives 6 months in Michigan and 6 months in Florida, they 
have emergency service needs and such. So they are arguing that 
they should get counted half and half.
    One of the problems--and in a way I wish you could have it 
on a form. If I have a place here in Washington, I fill out my 
form in my home here, and I fill out my form in Florida. If 
Members just throw the forms away it means that you are going 
to have to send an enumerator to knock on that door. I got my 
form in Washington, but it doesn't tell me what to do with it.
    Mr. Prewitt. Right.
    Mr. Miller. This is your second home. Longboat Key is a 
tourist area. It is a large mobile home park in my district.
    Mr. Prewitt. Obviously Longboat Key, the town, which is 
very low, it gets 50 percent, but half are seasonal homes. When 
we actually report the final number, which is different from 
the initial response rate, which is the return rate, it will 
come in at 100 percent. So they will get that credit, and we 
will make sure that they get that credit. And indeed across the 
country we know there to be roughly 9 percent or so of seasonal 
homes and vacant homes.
    Mr. Miller. How many?
    Mr. Prewitt. Nine percent of households or addresses in the 
United States, are one way or the other vacant.
    Now, I think your question, sir, on why we didn't have a 
better procedure in place for identifying the seasonal homes is 
a completely fair question. I wish we had. It would have been 
better to try to identify those households so we don't have to 
send out a nonresponse followup enumerator. Somebody will get 
to that neighborhood and say, ``yes, these five people have all 
driven up to Detroit,'' and they will be ticked off as seasonal 
and vacant housing units. In my judgment, if there was a better 
way, we should have done it.
    Mr. Miller. In Florida in the Tampa area, there were front 
page stories and concerns about problems within the Tampa 
operations. I am curious if you are aware of them and get your 
assurance that we are going to resolve them. I think the GAO 
has expressed that they would be willing to help out. I need to 
get your assurances that the problems in Hillsborough are going 
to be addressed?
    Mr. Prewitt. Well, two things if I could address there. 
First, the response rate right now from Hillsborough is within 
10 percent lower than its 1990 rate. There is not any kind of 
big variation from the response rate.
    Certainly in Tampa there is an early and continuing 
recruitment problem. That, sir, had to do with the quality of 
our management staff. We had to change the management staff, 
and we think that we have seriously upgraded it. I can't 
explain to you exactly what went wrong there today because the 
person who had to be let go has not signed his privacy release 
form, so we cannot discuss that. But the Census Bureau made the 
decision that we knew that we did not have strong management in 
the Tampa office, and we acted quickly and made sure that you 
do have strong management.
    We are expecting right now in the Tampa office not to hit 
our 100 percent recruitment goal. We expect by the time we 
close down the recruitment on April 19, we will be at about 70 
percent. However, we have already determined that in the 
surrounding areas we have an oversupply in our recruitment pool 
and that we will be able to borrow roughly the same kind of 
people that we would be hiring in Tampa. Once we put a good 
management team in place, the recruitment shot up. It was not 
that the labor pool was not there, our procedures were not 
effective.
    The Tampa article that you referred to, and I have in front 
of me, from the Tampa Tribune does use as its primary source of 
information the very individual that Carolyn Maloney just 
talked about. When a member of the Monitoring Board staff says, 
``Most cities say they are being road-blocked by the Census 
Bureau from completing their task,'' I would be hesitant to 
take that person's testimony as the testimony about what is 
going on. Who could actually believe that the Census Bureau is 
trying not to count cities across the country? He is 
attributing this to most cities in the country.
    So I would urge you not to over attribute a particular 
newspaper article, especially if the source of information is 
someone who is willing to make those kinds of charges.
    Mr. Miller. There are problems at Tampa, and so the problem 
is not just because one person made some statements that they 
obviously should not have made. They are legitimate problems, 
and you are addressing them, and the resources are there, and I 
think we can give assurances that everyone is going to do what 
they can.
    Mr. Prewitt. Not just because the subcommittee chairman 
happens to be from that area, but Tampa was one of the 
problems, and we did act aggressively and successfully, and I 
can be reassuring that we are now on schedule, on target. We 
will not hit our recruitment level, but we--don't forget, it is 
a 5 to 1 ratio, and so we don't need all of those people. 
Nevertheless we would have liked to have hit our target, but we 
are convinced that we have the number of people to do the 
nonresponse followup.
    Mr. Miller. Mrs. Maloney.
    Mrs. Maloney. Dr. Prewitt, for the sake of our television 
audience and people who may be watching this, what should 
someone do if they have not received their questionnaire and 
they would like to get their census form? What should they do?
    Mr. Prewitt. As I think the chairman correctly said, at 
this stage the most important thing to do is to call the 
telephone assistance number, the 1-800-471-9424 number, and we 
will still try to get the form to you. The reason that we 
stress that process is because by asking our system for a 
questionnaire, we then will have your address because we know 
where we have mailed it, which means that we can geocode it 
more easily when it comes back in.
    In addition, we have the Be Counted system, which is a 
safety net system. We hope that a lot of people don't have to 
rely on the Be Counted system because it is a much harder 
geocoding problem. We want people to use it if there is no 
other way.
    Finally, I do remind people there are certain people who do 
live in new construction, we will find them in new 
construction, and we also have the nonresponse followup. If 
there is an address, we will be knocking on the door if a form 
didn't come in.
    Mrs. Maloney. Again, for our listening public, if they 
received two forms, if they have two apartments in the same 
city or two houses so they have access to their other form, 
what should they do with the second form?
    Mr. Prewitt. If they have two separate residences, they 
have to follow the residency rules, which are problematic. We 
urge them to use the form at the residence that they most 
frequently occupy.
    Mrs. Maloney. And mail back the other one?
    Mr. Prewitt. That goes to the chairman's question. I got 
one at a place that I am not living, and I mailed it back in. I 
put in zero in terms of the number of people living there 
joping that we will get that out of our nonresponse followup. 
It will most likely be difficult to do that, of course, but 
maybe they will come in, and it will be a clue.
    Mrs. Maloney. Say someone has three apartments in one city, 
and they get three different forms. If they would mail back all 
three, would your system catch the name?
    Mr. Prewitt. We have a deduplication process, but in this 
case we do end up with an overcount, and one of the things that 
the accuracy and coverage evaluation does is identify the 
number of people, the proportion of people who end up sending 
more than one form in. In 1990, when we talk about the 
undercount number, we talk about a net. That is a difference 
between the number that we doublecounted and the undercounted. 
We try to find them and use the accuracy and coverage 
verification to detect that.
    Mrs. Maloney. I want to emphasize how unfortunate it is 
that talk show hosts have called the census long form optional. 
I want to compliment major newspapers and writers across this 
country that have come out with strong editorials in support of 
an accurate census and in support of the long form and urging 
everyone to not listen to any elected official who is saying 
otherwise. And I have with me the Seattle Times. We have Roll 
Call, Tulsa, the Washington Post, the New York Times, the 
Milwaukee Journal, the Atlanta Times, the Sacramento Bee, the 
Memphis paper in Tennessee, the Commercial Appeal, and they 
keep coming into my office, and so I think the press and the 
country has responded in a responsible way encouraging people 
to be part of this.
    I yield back the balance of my time.
    Mr. Miller. I have several other questions, but for the 
sake of time, we want to go on to GAO. I have some questions 
about proxy data and close-out verification. I would like to 
discuss that some more.
    Did you see the Dave Barry column the other day?
    Mr. Prewitt. Very funny.
    Mr. Miller. We have to have a sense of humor about this.
    Mr. Prewitt. No, I liked that one a lot.
    Mr. Miller. I know that you are very loyal about this, but 
you are missing your pin. Just sitting here--I know that you 
have dozens of them in every coat. You have been giving them 
away, but----
    Mr. Prewitt. I appreciate the chairman. Before we get off 
camera, let me get my pin on.
    Mr. Miller. Thank you again for being here. It is a tough 
job. I encourage everybody to complete the form. In conclusion, 
thank you very much, and I will see you next time.
    We ask Mr. Mihm, accompanied by Mr. Robert N. Goldenkoff 
and Mark Bird, to come forward, and I will swear you in.
    [Witnesses sworn.]
    Mr. Miller. Let the record note that they answered in the 
affirmative.
    Let me briefly say since we have people watching this that 
the General Accounting Office is a nonpartisan organization. 
They have a Web site that says the GAO's mission is to help 
Congress oversee Federal programs and operations to ensure 
accountability to the American people. GAO evaluators, lawyers, 
economists, public policy analysts, information technology 
specialists and other multidisciplinary professionals seek to 
enhance the effectiveness and credibility of the Federal 
Government.
    We rely on GAO for all of our congressional oversight. We 
appreciate them.
    Mr. Mihm, you were involved in the 1990 census, and so we 
appreciate the knowledge that you have contributed to this. At 
this stage let me ask you to make your opening statement.

 STATEMENT OF J. CHRISTOPHER MIHM, ACTING ASSOCIATE DIRECTOR, 
     FEDERAL MANAGEMENT AND WORKFORCE ISSUES, U.S. GENERAL 
ACCOUNTING OFFICE, ACCOMPANIED BY ROBERT N. GOLDENKOFF AND MARK 
              BIRD, U.S. GENERAL ACCOUNTING OFFICE

    Mr. Mihm. Thank you, Mr. Chairman and Mrs. Maloney. It is 
again an honor to appear before you today. I am joined by 
Robert Goldenkoff and my colleague Mark Bird, who has data 
processing responsibilities.
    This afternoon I will briefly hit the highlights of my 
written statement in six areas: first on the mail response 
rate; second on recruitment; third on update/leave operations; 
fourth, service-based enumeration or the counting of the 
homeless population; fifth on Questionnaire Assistance Centers; 
and sixth, data capture.
    First, in regards to the mail response rate, as Director 
Prewitt noted as of April 1, the national rate was about 55 
percent. Figures 1 and 2 in my prepared statement show the 
progress of the mail response at the regional and local levels. 
As you can see from those charts, overall the news is good thus 
far. Overall about 90 percent of Local Census Offices are 
three-quarters or more of the way toward achieving the final 
response rate they had in 1990, which, of course, is a higher 
benchmark than the Bureau has budgeted for. Meeting that would 
go a long way toward ensuring an accurate and complete census.
    Second, the Bureau is making progress in meeting its 
recruiting goals, but certainly continued efforts are still 
needed. As Director Prewitt has noted, the national goal of 2.2 
million qualified applicants has been met, but about 41 percent 
of the Local Census Offices have not met the March 30 
recruitment goal compared to about 53 percent that had not met 
the goal as of March 2. So we are seeing real progress at the 
national and local level, but we still have our 40 percent of 
the census offices that are not where they need to be in terms 
of recruitment.
    Third, over 24 million update/leave questionnaires were 
delivered by 70,000 census field staff. While national data are 
not yet available, our observations of update/leave suggest 
that update/leave made important improvements in the quality of 
the address list, including correcting for potential lapses in 
earlier address list development efforts. If these corrections 
are accurately reflected in the maps and address binders and 
keyed in accurately, they will reduce problems with nonresponse 
followup.
    Fourth, the Bureau's service-based enumeration operation 
attempts to count individuals who lack conventional housing 
when they go for services such as to shelters or soup kitchens, 
as well as attempting to capture them at targeted outdoor 
locations. Despite great effort on the part of the Bureau, the 
inherent challenge of counting this population combined with 
operational problems make the completeness and accuracy of this 
data uncertain. Overall, through several dozen field 
observations in 12 different locations, we noted that the 
operation was well staffed and received the cooperation of 
service providers. In addition, enumerators largely approached 
their jobs with professionalism and respect for the population.
    Mrs. Maloney, you mentioned that you and Chairman Miller 
were out in the streets and saw that firsthand. I had the 
opportunity to see it as well. For example, a team of 
enumerators I accompanied during the early morning hours of 
March 29, in Rosslyn, VA, searched heavy underbrush along the 
Potomac River. This was truly impressive. They searched under 
the walking bridge over to Roosevelt Island, there were three 
different ways they went in, and they were determined to find 
our encampment. They did find evidence that homeless people 
resided there, including the mattresses and clothes and other 
personal belongings.
    On the other hand, however, we also observed the challenges 
that the Bureau faces in trying to count individuals without 
usual residences. In some locations a police presence, the 
weather, the tornado down in Texas, and the terrain hampered 
enumerators' ability to find people living on the streets. In 
addition, however, a lack of sufficient supplies, inadequate 
enumerator training in some cases, inconsistent procedures for 
handling rejections and inadequate advanced planning undermined 
the quality of the count.
    Overall, while these problems may have affected the quality 
and completeness of the count and therefore should not be 
minimized, it is not surprising that they occurred in such a 
large and complex undertaking.
    My fifth point is that the Bureau continues to work to 
ensure that its 23,700 Questionaire Assistance Centers are 
available to the intended populations. My prepared statement 
provides examples from Laredo and Del Rio, TX, of some of the 
successful efforts that we observed. On the other hand, we saw 
less input from local partners and less promotion in other 
census offices that we visited in Oklahoma and Virginia, 
although assistance centers were open in those areas as well.
    Finally, data capture operations. As Director Prewitt 
pointed out, the data capture operations are working 
successfully. Available operational data tends to confirm that 
view. But some risks still remain that warrant continued 
attention.
    In our February report we expressed concern that the short 
time between the conclusion of the development and test 
activities of the data capture system and the date when data 
capture operations would begin created the risk that new 
problems would come to light after the system was in use. This, 
in fact, is occurring. In fixing these new problems, the Bureau 
has had to delay some important changes. As we discussed at the 
March 2 hearing, under the two-pass approach to data 
processing, the Bureau is making two sets of software 
modifications. The first set of changes were completed in 
February, and the second was to be completed by April 27. The 
Bureau has now delayed completion until May 31 because it needs 
to divert personnel to address the newly arising data capture 
problems. If new problems continue to surface, the completion 
of the second release will be increasingly at risk.
    On behalf of the subcommittee, Mr. Chairman and Mrs. 
Maloney, we will continue to track data processing and other 
key operations. This concludes my statement, and I would be 
happy to take any questions you may have.
    [The prepared statement of Mr. Mihm follows:]

    [GRAPHIC] [TIFF OMITTED] T0056.031
    
    [GRAPHIC] [TIFF OMITTED] T0056.032
    
    [GRAPHIC] [TIFF OMITTED] T0056.033
    
    [GRAPHIC] [TIFF OMITTED] T0056.034
    
    [GRAPHIC] [TIFF OMITTED] T0056.035
    
    [GRAPHIC] [TIFF OMITTED] T0056.036
    
    [GRAPHIC] [TIFF OMITTED] T0056.037
    
    [GRAPHIC] [TIFF OMITTED] T0056.038
    
    [GRAPHIC] [TIFF OMITTED] T0056.039
    
    [GRAPHIC] [TIFF OMITTED] T0056.040
    
    [GRAPHIC] [TIFF OMITTED] T0056.041
    
    [GRAPHIC] [TIFF OMITTED] T0056.042
    
    [GRAPHIC] [TIFF OMITTED] T0056.043
    
    [GRAPHIC] [TIFF OMITTED] T0056.044
    
    [GRAPHIC] [TIFF OMITTED] T0056.045
    
    [GRAPHIC] [TIFF OMITTED] T0056.046
    
    [GRAPHIC] [TIFF OMITTED] T0056.047
    
    [GRAPHIC] [TIFF OMITTED] T0056.048
    
    [GRAPHIC] [TIFF OMITTED] T0056.049
    
    Mr. Miller. The long form--one of the questions that I 
asked Director Prewitt was about the differential on the dress 
rehearsal, that was the 10 and 15 percent differential in the 
dress rehearsal. Director Prewitt didn't think that was a 
warning sign. Looking back at it, it should have told us there 
is a concern about privacy. It was too late at that stage to 
change the long form. We had to get the data, but maybe there 
was some other way we could have promoted it. Do you have a 
comment on that?
    Mr. Mihm. I think there were plenty of warning signs in 
hindsight, and that is why the Bureau sought to streamline the 
short and long form, make the entire approach more user-
friendly and have an advertising program that focuses on what 
the census means to you and your community. ``It is your 
future, don't leave it blank.''
    In addition to all of the issues that Mr. Ryan and you were 
mentioning, Mr. Chairman, there was a broad acknowledgment that 
generally public attitudes and concern about confidentiality 
and privacy and invasiveness were out there. In an electronic 
age those feelings are certainly strong.
    There was indeed a difference in the--or a growth in the 
difference in the long form/short form mail response rates 
between 1990 and the dress rehearsal. But on the other hand, as 
the Director has pointed out, mail response rates in the dress 
rehearsal are not predictive.
    One of the things that I need to take a look at is the 
differential long form/short form response rates from the 1988 
dress rehearsal before the 1990 census, and that will give us a 
feel whether or not there was more of an issue out there that 
we should have been attentive to.
    Mr. Miller. I would be glad if you would let us know.
    [The information referred to follows:]

    [GRAPHIC] [TIFF OMITTED] T0056.050
    
    [GRAPHIC] [TIFF OMITTED] T0056.051
    
    [GRAPHIC] [TIFF OMITTED] T0056.052
    
    [GRAPHIC] [TIFF OMITTED] T0056.053
    
    [GRAPHIC] [TIFF OMITTED] T0056.054
    
    Mr. Miller. Let me ask you about the data capture center, 
and I think the report is that things are going well. You 
mentioned that the Bureau assured you that the problems found 
in the four-site test have been resolved. Please discuss the 
problems experienced, and do you have documentation that the 
problems have been resolved?
    Mr. Mihm. The four-site test was the fundamental test that 
the Bureau did at the end of February, the 22nd to the 25th, 
that was to test all operations in an integrated way. In our 
testimony last time, we expressed some concern about the 
completeness of that test and the lack of information that was 
available to us at that point. We have since seen the report 
that has come out. Mark, you are most familiar with that.
    Mr. Bird. Yes. We received their report on the four-site 
test about a week ago, and we have reviewed it. The report 
itself does a good job of documenting many of the problems and 
the resolution of the problems. In addition, the system 
development contractor has a process for identifying, tracking 
and resolving problems, and that is an effective process.
    By way of example, one of the problems that was identified 
was that there was a discrepancy between the number of data 
files that had been transmitted to headquarters and the number 
of data files that had been reported as transmitted to 
headquarters.
    That discrepancy has been resolved.
    Mr. Miller. You mention that the contractor proposed 
eliminating system acceptance testing to ensure quality to save 
time. Please discuss that in further detail, and what are the 
implications in that?
    Mr. Bird. In a large system development and acquisition 
effort such as DCS 2000, it is important for the acquiring 
organization, which, of course, in this case is the Federal 
Government, to have some insight into the contractor's progress 
in the development of the system. Heretofore in the DCS 2000 
program, that has been accomplished in part by system 
acceptance testing, which has been witnessed by the government.
    So if, as has been proposed, system acceptance testing on 
the ongoing development work of DCS 2000 is eliminated, we 
would be concerned if there is no other opportunity for the 
government to witness testing. We don't yet know whether that 
is the case because the plans for the ongoing DCS 2000 
development have not been finalized.
    Mr. Miller. Let me ask about the recruiting, and I will 
bring up the Tampa issue. Recruiting can be successful in New 
York, but if you can't solve the problems in Tampa, there are 
surrounding areas that can fill in, I am assuming, in the St. 
Petersburg or Lakeland or some close-by areas.
    How serious of a problem is it? You said half of the local 
census service offices are understaffed at this stage as far as 
the number of potential nonresponse followup workers, and have 
they reacted adequately to address that issue?
    Mr. Mihm. About 41 percent have not met their most recent 
recruiting goal. This is a bit of an issue of concern. In a 
large national undertaking, a normal distribution applies. You 
have some that are doing very well and some that trail off at 
the end. And the national numbers showing success are taking 
advantage of the fact that the Denver and Dallas region are 
approaching 120 percent of the goal. And so it is a bit of a 
concern, or at least it is still a reason to continue to watch 
recruiting efforts--as Director Prewitt said, they certainly 
will continue to do aggressive recruiting down at the local 
level.
    In regards to your comment about how feasible is it to move 
people across areas and have them work in different offices, in 
some cases that can work. It adds additional travel cost, of 
course, to the Bureau because they do pay mileage for 
transportation. The issue, though, is that generally they find 
census takers want to enumerate neighborhoods that they are 
familiar with, and people want to be enumerated by people that 
they are familiar with. To the extent that you try and move 
people or ask people to work successfully in different 
neighborhoods, you usually find a lot of refusal, and you 
usually find that people are unwilling.
    Mr. Miller. How serious is that 41 percent that you are 
using; 41 percent of the LCOs are not adequately hired up?
    Mr. Mihm. It is hard to say at this point. They have 70,000 
people on the ground doing update/leave and didn't report 
significant staffing problems. As Director Prewitt noted, the 
big question is when they are going to have 500,000 enumerators 
on the ground doing nonresponse followup, and that becomes an 
enormous challenge for them. Thus far it appears that the 
recruitment program, the geographic pay rates that are higher 
and more aggressively managed than in 1990, and certainly the 
recruitment process generally is more aggressively managed than 
in 1990, seems to be paying off in many areas.
    But there are these 41 percent of the offices that, in our 
view, are the ones that bear some scrutiny. What we are going 
to be doing over the coming days as we get a better feel for 
where the mail response is shaking out for census offices is to 
compare these two and try to come up with a set of offices that 
are having both recruitment problems and mail response 
problems, and that will allow all of us--and I know the Bureau 
does the exact same thing--allow all of us to have a defined 
subset of what are the likely offices with the most challenge.
    Mr. Miller. Tampa had a management problem, and they don't 
necessarily correlate?
    Mr. Mihm. Not necessarily. In some cases they do. One of 
the things that I think is good to see this time is that the 
pattern from 1990. In 1990, they had a great number of problems 
with recruitment. In this--for 2000, you are still seeing some 
poor mail response. We discussed when Mr. Davis was here the 
problems that they have having in Chicago. They are having some 
problems in New Orleans, as well. There are 8 to 10 offices 
where they are having the biggest challenges in terms of mail 
response. Those are not necessarily the offices where they are 
having the biggest recruitment problems. In some cases there 
are correlations, but it is not as uniform as it was last time.
    Mr. Miller. Thank you.
    Mrs. Maloney.
    Mrs. Maloney. For the record, since it was such a large 
discussion at our last hearing, Mr. Mihm, have you had any 
access problems?
    Mr. Mihm. No, ma'am. On the contrary, I was able to talk to 
senior Bureau people over the last week, spoke with Director 
Prewitt and Deputy Director Barron today and told them that we 
continued to have good cooperation from them. Our access issues 
were resolved. We had a number of people that were on the field 
during the soup kitchen and shelter and the targeted nonshelter 
outdoor location, they were very, very cooperative and very 
accommodating. We are expecting that it will continue to be 
that way because of the efforts of the Bureau, and certainly 
the efforts of this subcommittee, to make sure that we had 
appropriate access.
    Mrs. Maloney. On the substance of your report, your 
testimony reflects the usual thorough job of GAO, and it points 
out a number of what I would call minor challenges, but it 
certainly doesn't seem to be anything that would threaten the 
success of the 2000 census. In fact, I read your testimony or 
hear your testimony as essentially good news. Is that a proper 
characterization?
    Mr. Mihm. I would agree, yes, ma'am. As we have been saying 
now for many months, the linchpin of a successful census is a 
high mail response rate. And at this point we are looking at a 
pretty good mail response rate. Depending on the bump that the 
Bureau gets over the next couple of days, the Census Bureau 
Director mentioned that they are at 57 percent, or that is the 
number that they will come out with today. Within the next day 
or so, we will see any bump that they got from April 1, and 
then if he gets another hit coming next week, we could be well 
over 61 percent and approaching the 1990 numbers.
    As we have said before, each percentage point is 1.2 
million fewer cases that need nonresponse followup and $34 
million that could be better spent.
    Mrs. Maloney. This is an important point that you raise. 
The two principal risks that you raised in December were the 
Bureau's mail response time and also the tight labor market 
which you have been discussing. Overall how would you rate the 
response rate? Very good? Extremely good?
    Mr. Mihm. At this point it does seem quite good. As I 
mentioned, 90 percent of the local census offices are at three-
quarters or more of the 1990 rate, which means that they are in 
striking range of the mail response rate that they got in 1990. 
I agree with what Director Prewitt was saying, that the 
relevant indicator for most district offices is not the 
national rate, it is doing better than you did in 1990.
    There are some areas of concern. The big issue now is--
irrespective of a good mail response rate--is to make sure that 
we get out of the field as quickly as possible. Even with the 
Bureau's assumptions, which would be a 61 percent mail response 
rate, they were still looking at following up on about 49 
million cases in 10 weeks, which is shorter than the amount of 
time than it took in 1990. So one of the concerns is as we get 
toward the end of this operation, are we closing out those crew 
leader districts, as the director mentioned, prematurely, or 
what kind of controls does the Bureau have in place that we do 
not go to last resort or proxy data before they should. That is 
the next big issue.
    Mrs. Maloney. Are you willing to make any predictions about 
where we might end up with these numbers?
    Mr. Mihm. I would prefer not. I wish I could. The Bureau is 
taking exactly the right position on this, and that is a tone 
of cautious optimism. They know, and their response model shows 
that as we get closer to that 61 percent and even closer to 65 
percent or 90 plus 5, it gets harder and harder to get, because 
there is a significant trail-off in mail response. In order to 
get to 61 percent, we are looking at basically another 750,000 
cases per day in each of the next 10 days. Can they make it? 
They certainly can, but on the other hand, I would not be 
necessarily shocked if we came in just right below that. But I 
think the news overall looks good for them on the mail response 
rate.
    Mrs. Maloney. How is the Bureau's Internet questionnaire 
progressing?
    Mr. Mihm. It had not been tested before, and it was not 
something that they put an enormous effort in. The Bureau had 
established the possibility of getting up to a million 
responses to that. The reality is quite a bit lower, and they 
are not necessarily disappointed with that. It is in the 
neighborhood of tens of thousands. It is about 60,000 or 
70,000.
    Mr. Goldenkoff. It is about 58,000.
    Mr. Mihm. We, at the request of the subcommittee, had done 
some preliminary looks at the security provisions that they had 
in place and came away convinced that, at least from the 
standpoint of the stated provisions, that they did have a 
secure system. They have done some testing to see if it could 
be hacked into. It has been successful in that regard.
    The big issue with the Internet is for the 2010 census. 
This came very late in the cycle and didn't get a dress 
rehearsal test. For 2010 we all need to look in a hard way at 
using the Internet, and technology generally needs to be 
seriously investigated, and I am sure the Bureau will do that.
    Mrs. Maloney. You commented that you felt the homeless 
organization could have been better organized. It certainly was 
not the experience that Mr. Miller and I had. They even swore 
us in. We said--they insisted on swearing us in, and we went 
out in a very organized way with the count.
    I read in the paper that Los Angeles, in that region they 
used individuals from the homeless community to accompany the 
enumerators as they went out on the street. Was that done in 
New York City? Was that a process that was followed across the 
country? It seems like a very good idea.
    Mr. Mihm. In regards to was it done in New York City, I am 
not sure. I do know it was a provision that the Bureau had 
nationally. Those people were technically called gatekeepers, 
and they were to be as you characterized, the representatives 
or very close or to even the homeless persons themselves that 
would basically be able to go into areas and say, the census is 
here, it is OK, it is important for us to be enumerated.
    In the observations that I did and my colleagues did, we 
didn't find that was necessarily the case that they used the 
gatekeepers. I didn't find, certainly in any of the 
observations that I did, it was a problem that those 
gatekeepers were not there. The census enumerators, as I 
mentioned in my statement, dealt with the people that they were 
enumerating with professionalism and respect for the dignity of 
those individuals. In fact, one of the mantras that the Bureau 
had is that we do not wake up people who were sleeping, and 
there were a number of people that I noticed, census 
enumerators, were waiting for people to wake up. Once they woke 
up, they would enumerate them. They made the correct judgment 
that it is better to have enumerators standing around rather 
than disturb someone that is asleep.
    Mrs. Maloney. Thank you very much.
    Mr. Miller. I have a couple more questions. There was an 
article in yesterday's CQ Daily Monitor about privacy on an 
appropriation subcommittee. There was somebody there from Eagle 
Forum, Public Citizen, from Public Interest Research Group, 
National Center for Victims of Crime and the ACLU.
    Privacy has become more and more of a concern. I think it 
is worth including this.
    You mentioned several problems in conducting the update/
leave operations. There are reports of children taking 
questionnaires off of doors or gates. What impact will all of 
these problems have on the quality of data from these regions? 
Should we be concerned?
    Mr. Mihm. Let me deal first with anyone removing a census 
form from a door. That would be then is presumably a 
nonresponse. It requires the Census Bureau to hire and train an 
enumerator to make up to the six visits to get that family in. 
That is a very unfortunate occurrence if it happens even one 
time, and extremely unfortunate if it happens quite often.
    The types of problems that we found were twofold. One is 
that the need to do extensive updating of the address 
registers, and the maps suggest in a positive way that doing 
update/leave was an important step in order to clean up those 
maps, and may have made some important additions and changes 
and improvements to previous address listing efforts.
    The key now will be to make sure that the changes get 
consistently included in the nonresponse packet. If an update/
leave enumerator went out there and found a problem with the 
map and corrected it, and that doesn't get corrected, then the 
census enumerator who goes out for nonresponse may have exactly 
the same problem. There should be a house here; I don't see 
that house. So there are some real efficiency concerns in both 
of those instances.
    Mr. Miller. I am hearing more and more counts of late or 
unavailable supplies and also the questionnaires in different 
languages, both from you and other field operations people. 
What is the reason for those problems, and how serious a 
problem is it?
    Mr. Mihm. We are still trying to find out the reason. The 
problem is across virtually all operations and across geography 
in the Nation. It does seem to be a nagging concern of a lack 
of supplies, and we are not just talking about the papers and 
pens, we have been focusing on training supplies not getting 
there in time. In the case of San Francisco, the short forms 
that they used to enumerate during the service-based 
enumeration did not get there in time, as I mentioned in my 
written statement, so they had to photocopy the forms, which 
requires that when the real forms come in, that they be 
recopied back at the local census office, because each has to 
have a unique identifier on them.
    There are a number of nagging stories of supplies not 
getting out, and whether it be training kits or foreign 
language recruitment material, the census in the schools not 
getting out in time, we are trying to still look at the causes 
of all of this. And it could be everything from it is in the 
local office and they don't know it yet--we have all been to 
some of these local census offices where we see boxes and boxes 
of material--to the distribution out of the Jeffersonville 
center. We are certainly going to be continuing to track the 
supply issue during nonresponse to see whether this is a 
pervasive problem.
    Mr. Miller. One important lesson learned from the dress 
rehearsals was the importance of clear expectations between the 
Census Bureau and community partners. It seems that the 
partnership program is having mixed results in 2000. Do you 
have a sense why this is occurring? Has the Census Bureau 
performed outreach uniformly across America?
    Mr. Mihm. They certainly offered. The 39,000 governments 
were offered the opportunity to participate. As we have 
reported in previous statements and in a couple of reports to 
the subcommittee, what we have found fairly consistently is a 
mismatch in expectations between local governments and the 
Census Bureau. Without going too far, it appears that a lot of 
this mismatch and expectations was particularly prevalent among 
some of the smaller or rural governments. Large cities have the 
expertise and experience to run a complete count type of 
program. They know what they are doing, and they understand 
clearly the stakes in an accurate count for them.
    The rural areas, especially when they have one or maybe 
even two employees at the local government, to ask them to take 
on the additional responsibilities of being the chief promoter 
and organizer of complete count in that community is onerous. 
They don't know how much they can rely on the Bureau. And so we 
have found some unevenness in the promotion and outreach 
campaign, particularly the complete count element of that.
    In order to get a more systematic view, and certainly to 
build for lessons learned, we are going to be doing some more 
detailed work down at the local level to try to get a feel both 
in successful areas and less successful areas asking what are 
the key ingredients of a profitable business partnership so we 
can build on that for 2010.
    Mr. Miller. Thank you.
    Mrs. Maloney. My last comment is that I hope everyone who 
has not filled out their form will be part of the census. Don't 
leave your future blank. This is a bipartisan effort. It is a 
responsibility of every resident in America, and as you pointed 
out, it is going to cost us more if you don't fill it out 
because we have to have enumerators. So it is important that 
you fill out your form.
    Mr. Miller. Thank you for being here. We appreciate GAO 
keeping on top of the issues.
    Next week I think we have the Congressional Monitoring 
Board before this subcommittee.
    I ask unanimous consent that all Members' and witnesses 
opening statements be included in the record. Without 
objection, so ordered.
    In case there are additional questions Members may have for 
our witnesses, I ask unanimous consent for the record to remain 
open for 2 weeks, and that the witnesses submit written answers 
as soon as practical.
    I would like to submit the Census Monitoring Board's 
congressional Members' request for oversight materials 
mentioned earlier for the record. Without objection, so 
ordered.
    The hearing is adjourned.
    [Whereupon, at 4:41 p.m., the subcommittee was adjourned.]
    [Additional information submitted for the hearing record 
follows:]

[GRAPHIC] [TIFF OMITTED] T0056.055

[GRAPHIC] [TIFF OMITTED] T0056.056

[GRAPHIC] [TIFF OMITTED] T0056.057

[GRAPHIC] [TIFF OMITTED] T0056.058

[GRAPHIC] [TIFF OMITTED] T0056.059

[GRAPHIC] [TIFF OMITTED] T0056.060

[GRAPHIC] [TIFF OMITTED] T0056.061