[House Hearing, 106 Congress]
[From the U.S. Government Publishing Office]
OVERSIGHT OF THE 2000 CENSUS: STATUS OF NON-RESPONSE FOLLOW-UP
=======================================================================
HEARING
before the
SUBCOMMITTEE ON THE CENSUS
of the
COMMITTEE ON
GOVERNMENT REFORM
HOUSE OF REPRESENTATIVES
ONE HUNDRED SIXTH CONGRESS
SECOND SESSION
__________
MAY 5, 2000
__________
Serial No. 106-191
__________
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
70-439 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
David McMillen, Minority Professional Staff Member
C O N T E N T S
----------
Page
Hearing held on May 5, 2000...................................... 1
Statement of:
Prewitt, Kenneth, Director, Bureau of the Census; John H.
Thompson, Associate Director for Decennial Census; and
Marvin D. Raines, Associate Director for Field Operations.. 73
Letters, statements, etc., submitted for the record by:
Maloney, Hon. Carolyn B., a Representative in Congress from
the State of New York:
Article dated May 4, 2000................................ 17
Prepared statement of.................................... 22
Miller, Hon. Dan, a Representative in Congress from the State
of Florida:
Congressman Tom Davis radio address...................... 3
Executive state of the Census report..................... 9
Information concerning invasion of privacy............... 7
Prepared statement of.................................... 11
Prewitt, Kenneth, Director, Bureau of the Census:
Letter dated July 6, 2000................................ 84
Prepared statement of.................................... 32
OVERSIGHT OF THE 2000 CENSUS: STATUS OF NON-RESPONSE FOLLOW-UP
----------
FRIDAY, MAY 5, 2000
House of Representatives,
Subcommittee on the Census,
Committee on Government Reform,
Washington, DC.
The subcommittee met, pursuant to notice, at 10 a.m., in
room 2247, Rayburn House Office Building, Hon. Dan Miller
(chairman of the subcommittee) presiding.
Present: Representatives Miller and Maloney.
Staff present: Jane Cobb, staff director; Chip Walker,
deputy staff director; Lara Chamberlain and Amy Althoff,
professional staff members; Mike Miguel, senior data analyst;
Andrew Kavaliunas, clerk; David McMillen and Mark Stephenson,
minority professional staff members; and Earley Green, minority
assistant clerk.
Mr. Miller. Good morning. Welcome to the May hearing with
Director Prewitt on the status of the decennial census. We will
begin with opening statements, and then we will have a chance
for Mrs. Maloney and myself to ask some questions of Director
Prewitt.
Thank you, Director Prewitt, for once again being here.
Since we last met, the Census Bureau has reported on the final
numbers for the mail response rates. The final mail response
rate of 65 percent will be at least 4 percentage points above
what the Bureau had budgeted for. As you have said, Director
Prewitt, this was no small achievement. The mail response rate
had been in steady decline since 1970. In the absence of
significant improvements, the mail response rate would have
been expected to be in the neighborhood of 55 percent this
time.
The Census Bureau is to be commended for halting the slide
in civic participation in the mail out/mail back phase of the
census.
I firmly believe that the combination of community
partnerships, paid advertising and a strong commitment to the
census by Congress--which in the end will have appropriated
almost $6.8 billion, have all contributed to the better than
expected mail response rate. A story in yesterday's New York
Times reported that all signs seem to indicate that the
outreach advertising and partnership programs have succeeded in
raising the response rates for those missed in the 1990 census
or at least preventing them from declining. This is significant
since Republicans have maintained that if we funded the proper
outreach and promotion programs, we could reach the
undercounted. I'm gratified to see we were right.
I am, though, still disappointed that three significant
programs were not included in this census. A second mailing,
which easily could have boosted response percentage rates into
the 70's, based on the results of the address rehearsal, the
use of administrative records and the ability of local
governments to check the Census Bureau's work. In fact, on the
final issue of post census local review, a local government in
the Tampa area has already decided to sue the Census Bureau.
Director Prewitt, in a letter dated April 14, I asked that
you reprogram the budgetary savings from an increased mail
response rate to reach those groups that are traditionally
undercounted. In that letter I estimated the savings to be
about $34 million for every percentage point above 61 percent.
This estimate was based on a report issued by the General
Accounting Office in December 1999. I also explained that I
would be of any assistance in gaining approval from the
Congress to transfer money between frame works. To date my help
has not been solicited. And in a written response to me you
also noted that although you believed there would be budgetary
savings, you believe that the GAO estimate may not be accurate
because of a lower than expected enumerator productivity rate.
Fair enough. I want to be clear on one point. This chairman
and this Congress expect you to use all of the tools in your
tool box to reach the undercounted. This windfall in your
budget is expected to be used directly to reach those not
counted during the mail response phase of the census and those
traditionally undercounted. This opportunity must not go to
waste. It would not be acceptable to miss our objectives and
have funding left to spare. More advertising, more outreach,
higher pay rates and special enumeration techniques must be
considered to help eliminate the differential undercount during
the most difficult part of the full enumeration, non-response
follow-up.
And speaking of non-response follow-up, I was delighted
that the House leadership devoted part of the Republican radio
address on April 22 delivered by Congressman Tom Davis, a
member of this committee, saying,
Next week, hundreds of thousands of enumerators will fan
out across the country to find those not already counted. These
enumerators are your neighbors and friends, co-workers and
family. When an enumerator comes to your door, please cooperate
by giving them a few minutes of your time and answering their
questions. By law your answers are kept strictly confidential.
Your census answers are important to allocate seats in Congress
and to help government officials determine where to build
roads, day-care facilities and schools. In the upcoming weeks,
if you should encounter a census worker, please thank them for
their effort and dedication to the 2000 census.
I want to personally thank Congressman Davis for delivering
this important message.
[The information referred to follows:]
[GRAPHIC] [TIFF OMITTED] T0439.001
[GRAPHIC] [TIFF OMITTED] T0439.002
[GRAPHIC] [TIFF OMITTED] T0439.003
Mr. Miller. Director Prewitt, there remains a great deal of
debate surrounding the long form. This subcommittee has been
trying to get a handle on just what is fueling this debate. Is
there really a legitimate feeling out in the public that the
long form questions are intrusive? Or, as some have charged, is
this debate being fueled by a few elected officials who have
expressed concerns for their constituents' privacy worries?
Dr. Prewitt, when you came before the subcommittee about a
month ago, and in numerous public events since, you cited a
poll by InterSurvey. You have claimed that people's uneasiness
about the long form jumped the week congressional leaders made
their remarks. What you neglected to say was that in fact the
bump in concerns coincided with the arrival of census
questionnaires in people's homes.
When we went back and looked at the polling data, it shows
that the rate of concern had actually reached 18 percent by
March 26--before the comments by Senator Lott and Governor Bush
were widely reported in the press. The reason why the previous
surveys showed the lower levels of concerns was because the
forms had yet to be mailed. What's more, the very next week
after what was supposedly alarming remarks, the concern rate
over the long form fell 2 percentage points. I am very
disappointed that you were not more forthright regarding this
poll which is being conducted in conjunction with the Census
Bureau. Since April 18th, you have known that your worries
about the long form have been ``resolved,'' and that long and
short form return rates have exceeded your expectations. Yet
you have continued to express concerns about the long form and
blame Republicans for their comments.
[The information referred to follows:]
[GRAPHIC] [TIFF OMITTED] T0439.004
Mr. Miller. I can only conclude that since your public
comments do not match your own internal information, you are
attempting to politicize the census at this crucial period of
time.
Director Prewitt, let me call your attention to the next
chart. This is a copy of page 5 of the April 18 Executive State
of the Census report produced by the Census Bureau. It clearly
states that issue regarding the long form response had been
resolved.
Resolved, Long Form Response Rate--The difference between
the response rate for the long form and short form has been
greater than expected. We were concerned because conducting
proportionally more long form interviews affect productivity in
non-response follow-up. Resolution: By April 18, both the long
form return rate and short form return rate have exceeded our
goals.
[The information referred to follows:]
[GRAPHIC] [TIFF OMITTED] T0439.005
Mr. Miller. While internally this issue is, ``unresolved,''
you have continued to overstate the problem. I have to say that
I am disappointed that the head of an agency that prides itself
on accuracy and quality of data would succumb to these
political temptations. At the same time, I realize that you are
a political appointee of President Clinton, and as such, are
subject to the influences of this administration.
As I have said before, this administration is as much to
blame for these increasing privacy concerns as anyone is. From
the Pentagon to the White House, this administration has
demonstrated time and time again that it only believes in
privacy when it is politically expedient. President Clinton and
Vice President Gore must be paying attention to the current
privacy issues regarding the long form because they have just
launched a new privacy initiative. I find this almost laughable
considering the breaches of trust this administration has been
accustomed to.
Let me also say how deeply concerned I am about the
accidental faxing of confidential information to a private
household that recently occurred in Congressman Coburn's
district. For our viewing and listening audience let me give
some of the facts as reported in the Phoenix newspaper earlier
this week.
A Census Bureau employee at the regional office
accidentally dialed in a wrong fax number and faxed information
on Census Bureau applicants to a private household instead of
another census office. This information included names,
addresses, test scores and Social Security numbers and is
protected by the Privacy Act. The fax was then given to
Congressman Coburn and that is how this serious breach of
security, even if accidental, came to light.
I have been a staunch defender of the Bureau's commitment
to privacy, but frankly that confidence has been shaken. You
cannot placate Members of Congress and the American people who
have expressed concerns about privacy and confidentiality on
the one hand and then allow this kind of thing to happen on the
other. I certainly can't assure people with the same level of
confidence I had a week ago about the Bureau's ability to
protect their privacy.
Director Prewitt, the Founding Fathers were very wise. I
now know that the real reason we only conduct the census every
10 years is because no one can possibly go through this process
yearly, whether on your side or mine. This has been truly an
arduous task, but it is made more difficult when we see a
pattern of behavior that lends itself to partisan politics. You
made a commitment to be nonpartisan, and I will hold you to it.
Mrs. Maloney.
[The prepared statement of Hon. Dan Miller follows:]
[GRAPHIC] [TIFF OMITTED] T0439.006
[GRAPHIC] [TIFF OMITTED] T0439.007
[GRAPHIC] [TIFF OMITTED] T0439.008
[GRAPHIC] [TIFF OMITTED] T0439.009
[GRAPHIC] [TIFF OMITTED] T0439.010
Mrs. Maloney. Thank you, Mr. Chairman.
On April 27th, the most critical and labor intensive phase
of census 2000 began, as census takers fanned out across
America to visit those households which did not mail back their
questionnaires. These next 10 weeks will undoubtedly be the
most difficult faced by the Bureau during the 2000 census.
I urge all Americans to cooperate with these census
takers--people from their own communities who have undergone a
security screening and who will be easily identifiable.
For the most part, these workers are your neighbors and
friends, hired from the local community because they know its
streets and neighborhoods, speak its languages, and are
familiar with its cultures.
Your cooperation is vital to the success of the 2000
census. Your answers are strictly confidential. No other
government agency or private individual will see your answers--
not the IRS, the FBI, the INS, or the CIA. Please cooperate if
an enumerator knocks on your door.
When you look back only a few months, the two biggest
unanswered questions that had the potential to threaten the
success of the census were what would the mail response rate be
and would we be able to hire enough qualified workers to do
non-response follow-up in the midst of this incredible economy?
Well, we now have the answer to both of these questions.
First, the Census Bureau through its remarkable advertising
campaign and community outreach efforts has reached a 66
percent mail response rate for the 2000 census, an outstanding
achievement which has reversed the decades-long decline in the
participation of the American people with the census.
Second, as a result of careful planning, the Bureau has
recruited 108 percent of its national hiring goal and I must
say, Director, that having met many of these enumerators while
working with the chairman on homeless night here in the
district and while visiting with workers in Queens and
Manhattan, I am really impressed with the people you have
recruited.
The commitment and energy that they show to the task of
counting America is inspiring given what I know is a
challenging job of knocking on doors and trying to get people,
especially New Yorkers, to take a minute and talk to you.
I would like, Mr. Chairman, to place in the record an
article from yesterday's Boston Globe written by an enumerator
that I think captures the spirit shown by enumerators. How hard
they are working and how dedicated they are to their job, both
counting Americans and keeping the information strictly
confidential.
[The information referred to follows:]
[GRAPHIC] [TIFF OMITTED] T0439.011
[GRAPHIC] [TIFF OMITTED] T0439.012
[GRAPHIC] [TIFF OMITTED] T0439.013
Mrs. Maloney. These accomplishments are truly good news,
and I must commend Dr. Prewitt, Marvin Raines, John Thompson
and the entire decennial staff, and every employee of the
Census Bureau, both permanent and temporary, for a job well
done so far.
It appears that the census is on track.
Obviously in any operation as large as this there are going
to be problems, problems that I am sure the chairman's
questions will bring out in detail. But to me it seems that you
and the staff have tried to meet these challenges head on, that
you have been quick to inform the chairman and the public of
the problems, something I don't think many organizations would
do in such a quick and complete manner. You have warned us of
what you think the challenges will be.
So while the news nationally is indeed good, it still means
that a lot of work needs to be done, over a third of America's
households must still receive a visit from the census taker.
That's 42 million doors that need to be knocked on. I look
forward to hearing from you, Dr. Prewitt, on how the next
aspect of the census is coming and what we can expect.
But I really want to respond to some of the statements that
Chairman Miller just made. I only want to speak for myself. I
have never said that statements of Governor Bush, Senator Lott
and Speaker Hastert, along with a dozen additional Members of
Congress, are solely to blame for the privacy issues which have
been raised about the census. But I must say, and I think the
facts are very clear and speak for themselves, that the
leadership of the Republican Party in the middle of a national
civic ceremony in a national effort to count every single
person in our country, to get vital information about our
country so that we can plan and distribute Federal dollars
fairly, they decided in the midst of this campaign to count
everyone to go negative. They decided that they would not
support this national effort but would trash it. They didn't
show leadership and they didn't explain that all of this
information is completely protected. What they did was pander
to talk shows and right wing fringe groups. What they have
done--and I would like to put in this record what they have
done--in the midst of this is send out fundraising appeals
calling it the Republican census document. That is what their
effort is in the middle of this national civic ceremony.
I really believe very strongly that privacy is a
tremendously important issue to every person in America, and I
feel strongly about privacy and along with the leadership on
the Banking Committee in a bipartisan way, Chairman Leach and
many Democrats, and I was part of that effort, worked to put
forward privacy language in the banking modernization bill.
The President has come forward with even more language on
protection of financial information and he has put that before
Congress and I will be a cosponsor of it. On another committee
that I work on, Chairman Burton's committee, there have been
many, many hearings on privacy over health records, and in a
bipartisan way working with Chairman Horn, we have had many
hearings and put forward legislation and worked for privacy in
a bipartisan way.
But the census is protected. The confidentiality is
protected and it is important for planning for our country, and
as we have said many, many times, the questions on the census
form are exactly the same questions on the long form that
President Reagan and President Bush and every Member of
Congress that got 3 years notice endorsed. It is even shorter
than the form that went out in 1990. The only new question was
added in response, as we know, to the welfare reform in a
bipartisan way to get a tracking of how many grandparents are
taking care of children. So I must say that the timing of the
national Republican leadership in the midst of the most
sensitive time during the mail back response time to basically
call the census optional was just plain wrong.
May I put in the record the Republican national--it is
called the Republican Census Document.
[The prepared statement of Hon. Carolyn B. Maloney and the
information referred to follow:]
[GRAPHIC] [TIFF OMITTED] T0439.021
[GRAPHIC] [TIFF OMITTED] T0439.022
[GRAPHIC] [TIFF OMITTED] T0439.014
[GRAPHIC] [TIFF OMITTED] T0439.015
[GRAPHIC] [TIFF OMITTED] T0439.016
[GRAPHIC] [TIFF OMITTED] T0439.017
[GRAPHIC] [TIFF OMITTED] T0439.018
[GRAPHIC] [TIFF OMITTED] T0439.019
[GRAPHIC] [TIFF OMITTED] T0439.020
Mr. Miller. Without objection.
Mrs. Maloney. Thank you.
Mr. Miller. Well, I must take the chairman's prerogative to
respond briefly to this. I am very disappointed. To say that
the Republican leadership trashed the census, that is extreme
political rhetoric. I am extremely disappointed. ``Trashed the
census?'' Mrs. Maloney, the Speaker had a press conference with
me a couple of weeks ago. We took time on the Saturday radio
response to talk about the census. We provided every penny the
Bureau has asked for. They may have been given more money. And
to say that we trashed it is wrong.
When Members of Congress are responding to concerns of
constituents, that is what Members of Congress are supposed to
do. And then when we talk about this letter that was sent out,
we have invested over $7 billion in the census and I take my
role very seriously, and I try not to interject partisan
politics in the process. So when the Southeastern Legal
Foundation mailing went out, I put aside the fact that this
group was responsible for a major ruling by the Supreme Court
regarding the census. Their mailing did cross the line and I
said so.
I didn't stick my head in the sand and blindly defend them,
but any person looking at a mailing from the Republican
National Committee talking about a Republican, unless you want
to have a bill banning everybody from using the word census or
Census Bureau, this is clearly--it took the Postal Service less
than a day or so to say there was no rule broken. This is a
frivolous claim made in an obvious attempt to score political
points, and I would like to call upon my colleagues to join
with me in stopping to play politics.
Director Prewitt, would you and Mr. Raines----
Mrs. Maloney. May I respond?
Mr. Miller. Let's get moving with the opening statements.
Director Prewitt, would you rise, and Mr. Thompson and Mr.
Raines.
[Witnesses sworn.]
Mr. Miller. Let the record note that they have answered in
the affirmative. We appreciate that all of you are here again
today.
Director Prewitt, next week you have been asked to serve
jury duty, and there can't be a busier person in America right
now in the middle of the census than the Director of the Census
Bureau, but as we have all talked about the civic
responsibility of the census, it is a civic responsibility to
serve our communities on jury duty. So I commend your
willingness to step aside from your responsibilities as
Director so you can serve on the jury, and thank you.
Thank you once again for being here and you have an opening
statement. The official statement of course will be entered in
the record.
[The prepared statement of Mr. Prewitt follows:]
[GRAPHIC] [TIFF OMITTED] T0439.023
[GRAPHIC] [TIFF OMITTED] T0439.024
[GRAPHIC] [TIFF OMITTED] T0439.025
[GRAPHIC] [TIFF OMITTED] T0439.026
[GRAPHIC] [TIFF OMITTED] T0439.027
[GRAPHIC] [TIFF OMITTED] T0439.028
[GRAPHIC] [TIFF OMITTED] T0439.029
[GRAPHIC] [TIFF OMITTED] T0439.030
[GRAPHIC] [TIFF OMITTED] T0439.031
[GRAPHIC] [TIFF OMITTED] T0439.032
[GRAPHIC] [TIFF OMITTED] T0439.033
[GRAPHIC] [TIFF OMITTED] T0439.034
[GRAPHIC] [TIFF OMITTED] T0439.035
[GRAPHIC] [TIFF OMITTED] T0439.036
[GRAPHIC] [TIFF OMITTED] T0439.037
[GRAPHIC] [TIFF OMITTED] T0439.038
[GRAPHIC] [TIFF OMITTED] T0439.039
[GRAPHIC] [TIFF OMITTED] T0439.040
[GRAPHIC] [TIFF OMITTED] T0439.041
[GRAPHIC] [TIFF OMITTED] T0439.042
[GRAPHIC] [TIFF OMITTED] T0439.043
[GRAPHIC] [TIFF OMITTED] T0439.044
[GRAPHIC] [TIFF OMITTED] T0439.045
[GRAPHIC] [TIFF OMITTED] T0439.046
[GRAPHIC] [TIFF OMITTED] T0439.047
[GRAPHIC] [TIFF OMITTED] T0439.048
[GRAPHIC] [TIFF OMITTED] T0439.049
[GRAPHIC] [TIFF OMITTED] T0439.050
[GRAPHIC] [TIFF OMITTED] T0439.051
[GRAPHIC] [TIFF OMITTED] T0439.052
[GRAPHIC] [TIFF OMITTED] T0439.053
[GRAPHIC] [TIFF OMITTED] T0439.054
[GRAPHIC] [TIFF OMITTED] T0439.055
[GRAPHIC] [TIFF OMITTED] T0439.056
[GRAPHIC] [TIFF OMITTED] T0439.057
[GRAPHIC] [TIFF OMITTED] T0439.058
[GRAPHIC] [TIFF OMITTED] T0439.059
[GRAPHIC] [TIFF OMITTED] T0439.060
[GRAPHIC] [TIFF OMITTED] T0439.061
[GRAPHIC] [TIFF OMITTED] T0439.062
[GRAPHIC] [TIFF OMITTED] T0439.063
STATEMENTS OF KENNETH PREWITT, DIRECTOR, BUREAU OF THE CENSUS;
JOHN H. THOMPSON, ASSOCIATE DIRECTOR FOR DECENNIAL CENSUS; AND
MARVIN D. RAINES, ASSOCIATE DIRECTOR FOR FIELD OPERATIONS
Mr. Prewitt. If I may read a very quick opening statement
and take an extra minute or two to address some of the
questions that you raised in your opening statement.
Mr. Chairman, Mrs. Maloney, members of the subcommittee, I
am pleased to be here today to provide an update on the status
of the census activities. Last week I had the honor to report
the good news about the state of civic responsibility in our
country. The country has stopped a 30-year decline in census
cooperation, slightly reversed the decline, and this is a
serious achievement.
In reaching the 66 percent mail back response rate, the
public outperformed expectations. More than 100,000 census
partners deserve credit. Congratulations are owed to thousands
of mayors, commissioners, teachers, community advocates, houses
of worship and other civic business leaders. We thank our
partner agencies for the excellent advertising campaign and to
you, Mr. Chairman, Mrs. Maloney, other Members of Congress, who
encouraged response to the census. Our partners and the public
have treated the census as a serious civic event intended by
the founders.
The good news about the mail response rate is tempered by
our concerns about long form noncooperation and potential loss
of data. As I explained at the last hearing, every question we
asked in the census serves an important purpose and all have a
specific Federal or judicial mandate or requirement. Very early
this year an advocacy group issued a press release that said as
follows: ``real Americans don't answer nosy census questions.
You can strike a blow for privacy, equality and liberty by
refusing to answer every question on the census form except the
one required by the Constitution: How many people live in your
home?'' This is a misreading of the Constitution, which states
that the census is to be conducted ``in such manner as
[Congress] shall by law direct.'' The mistaken reading of the
Constitution ignores the fact that the Nation's founders and
its first Congress directed the tabulation of the population by
such characteristics as age, gender, race and household
composition. Every census has been more than a simple head
count. Moreover, the misguided advice on how to respond to the
census is a prescription not only for poor data quality but for
increased undercount.
If people do not cooperate with the census at all, or just
give us a number of persons in the household, whether when they
return the form by mail or when the enumerator visits, that
will not be sufficient. Beyond the number of people at an
address, we require some minimal characteristics to complete an
enumeration. Otherwise we have no way to know whether we are
dealing with real people. In cases where no cooperation is
forthcoming, we will have to attempt to get the data through
interviews with other knowledgeable individuals.
We are also concerned about potential loss of data due to
opposition to the long form. There was approximately a 12
percentage point difference between the mail response rates for
the long form and short form, double the 1990 rate. We do not
have data at this point about item nonresponse rates. That is,
for example, how many people who mailed back the long form did
not answer specific questions, such as income, disability,
education and so on. Comments we have received give us reason
to be concerned about the long form problem.
Let me cite just two of these comments. ``I have this day
read my long form and promptly ripped it in two and burned
same. Don't bother sending another as I won't fill it out nor
will I pay the $100 fine.''
Second, ``I am refusing to complete the long form. You can
arrest me if you want, but I am not going to complete it.''
Obviously this is a very small sample from a large number. We
are very concerned that refusal to respond fully to the census
can pose a serious risk to census 2000 data. As I previously
testified, the Census Bureau would have to determine whether
the data are sufficiently reliable to perform the functions
expected of them.
Let me turn to an operational update. In each of the
hearings that have tracked census operations, I have identified
problems that could put the census at risk in the period
following the hearing. Thus in the last hearing I listed as
potential problems the failure to complete the update leave
operation, problems with our payroll system, widespread
problems filling our enumerator positions, problems with the
address file, breakdown of our telephone questionnaire
assistance operation, breakdown of data capture, questionnaire
delivery and unexpectedly low mail response rates or any event
such as a hacker on our Internet site. None of those potential
problems has occurred. Every major census operation scheduled
for completion is either now complete or in its final stages.
This includes update leave, remote Alaska, service based
enumeration, military enumeration, foreign language
questionnaires and others. And I can provide details if you
wish.
Now we, of course, enter the nonresponse followup
operation, which is the largest, most complex and most costly
operation of census 2000. It raises its own set of potential
risks, and I take this hearing as an opportunity to put those
on the record. These would include high turnover rates for
enumerators, more outright resistance from respondents that
could affect productivity or data quality, a breakdown in our
payroll system or random events such as attacks on enumerators
or natural disasters. Turnover has been very low in early
census operations such as update/leave, but nonresponse
followup is a more difficult and frustrating operation. The
controversy over the long form, as I have said, gives us some
reason to be concerned about resistance and data quality. Our
payroll system has worked very well so far, but nonresponse
followup is such a big operation that it will be a major test
for that system. So, we face potential risks during nonresponse
followup that could affect accuracy, data quality and budget.
I want to emphasize that the Census Bureau will fully apply
its procedures to account for every address that is on our list
to be visited during nonresponse followup. These procedures are
extensive and include making up to six attempts, three by
personal visit and three by phone when a phone number is
available, to complete the enumeration of a household. These
procedures also include extensive quality assurance procedures
and supervisory controls but they also reflect our experience
that the longer we are in the field and the farther we get from
census day the more the quality of respondent's answer
deteriorates.
It is important to keep in mind that we are using a part-
time temporary staff to which we have been able to provide only
basic training in survey methods. Extending nonresponse
followup beyond the already extensive level of effort we plan
would not only increase census cost but it could lead to a
reduction in data quality. Mr. Chairman, we appreciate your
letter with respect to directing the resources obviously to the
hard to enumerate areas, and that is what we are doing. I have
not directly responded to you on a framework reprogramming. It
is not a framework issue at this stage, but certainly we are
putting the money in those areas. We have raised enumerator pay
rates in about 10 percent of our LCOs, including Tampa.
The preparation for and launching of nonresponse followup
is very time sensitive and it had to be completed in a few days
so we could begin training on time. While it was going on, we
continued to receive mail responses. Some of those made it into
our late mail return files, but some did not. Some people who
have mailed back their form will be visited in nonresponse
followup. We realize this will irritate some members of the
public who will wonder why we are bothering them again. There
had to be a cutoff date to begin preparing the assignments and
to get all of the maps and kits to the right training sites. We
do the best we can to strike the late forms that come in to the
nonresponse followup universe, but clearly cannot do so for all
late returns. Forms are still coming in. We have also received
be counted forms that do not have identification codes. These
require a labor intensive matching and place coding operation
to code them to the right geographic area. So this sometimes
correct complaint that I have already sent the form in is
something our enumerators are trained to deal with. Of course
they will try to complete an enumeration at these housing units
anyway because many will say that they have returned a
questionnaire even when they haven't.
We have sufficient staff to begin nonresponse followup on
schedule in every census office in the country. We have
frontloaded our training selections, which means that our goal
is to train and give assignments to twice as many people as we
needed. That way, we will have staff to offset attrition. We
have identified over 50,000 individuals for replacement
training so we can keep replenishing the pool of available
workers. We have retained this 2 to 1 redundancy at the vast
majority of sites. Across the national system, we are at 3 to 1
redundancy. We have 3 times the number of enumerators already
hired. So that simply means we have more people out there, and
we will have the opportunity to accelerate the completion in as
many LCOs as possible.
Nevertheless we continue to recruit in targeted areas even
as we speak. This may mean in the end that some qualified job
applicants may not be hired. We realize they will be
disappointed, but we believe we must keep the applicant pool
active to assure we have sufficient staff to cover attrition.
Thus far, we have identified 2.6 million qualified applicants
or 108 percent of our goal.
To place nonresponse followup in context, appendix 1
graphically depicts each of the major census enumeration
operations that precede and follow it. On these operational
issues, I will take your questions.
May I ask for a few moments to address the question that
you raised about the politicization of the conversation about
long and short form. Let me first bring to your attention
what--the second of your graphs, it was the census report that
you referred to and just interpret that so you will see what
that means.
What that ESOC report of April 18 reported was that based
on our nonresponse followup workload, the fact that we received
at that time a more than 4 percent increase over our expected
mail back response rate, meant that we were now convinced that
completing nonresponse followup on schedule was not at risk.
That is all that meant. We didn't resolve any issue about the
long and short form differential. It meant in terms of our
overall response, it was above the level that we needed to set.
It says nothing about data quality and completeness of the long
form data. We may well have a data quality problem but we
simply don't know that yet.
So it is disingenuous to say that we have resolved the
problem. We don't know. We have resolved the problem of
nonresponse followup as best as we can at this stage.
Let me turn to the other concerns that you expressed and I
appreciate the seriousness of them and I would like to take a
moment to address them. First, I have to say that perhaps it is
an accident or perhaps it is not an accident, that nothing in
your prepared comments that you just read from quote me as
calling into question the leadership of the Republican Party.
There is no quote available to have put into these comments
because my comments have never addressed the role of the
Republican leadership; and, therefore, I have to express some
concern that you have chosen to interpret my public comments as
chastising or otherwise criticizing the Republican leadership.
If I have, I ask you for that quote, whether it was in a press
conference or report or testimony. I don't believe such a quote
exists. There may have been newspaper articles that have
implied that, but that is not what I have said. Because I don't
believe that I have said that, sir.
I want to say what I have said publicly. What I have said
is that national public voices, which certainly includes some
of the leading members who have control over the airwaves, talk
show hosts, 60 Minutes, have undermined, as far as I am
concerned, the seriousness of the census and they did so during
a key period, and whether that is the third or fourth week of
the census is not the moment I was addressing. I was addressing
the moment that this conversation began to occur publicly.
I have also said, and here I have referenced national
political leaders, not just public voices, I have said that at
a key moment in the census, approximately March 27 to April 2
or 3, we had the full attention of the American people, the
full attention of the American people on the census. This is a
remarkable accomplishment. All of our information on exposure
and awareness suggest that 97, 98, 99 percent of the American
people were aware of the census. I believe that was a moment
when we could have had an important conversation with the
American public about the fact that democracy has to do with
rights and responsibilities as well as benefits. I believe we
missed that opportunity. I believe in that key week that what
could have happened--we could have said look, the census is
part of the responsibility of belonging to this country. And
that was not a good moment to talk about the census as a pick-
and-choose opportunity. If you don't like it, don't worry about
fully cooperating. That was not a good moment for those voices
to be heard.
My concern and what I expressed in public shortly after
that was out of the disappointment of a bipartisan passed
Senate resolution on the floor which subsequently was removed
in committee, and I appreciate the efforts that went into
removing that from the committee, but the floor nevertheless in
a bipartisan vote said, well, the census after all could be
thought of as a form of harassment, these enumerators knocking
on your door, that it is not something that should be
mandatory.
The reason that the census is mandatory, it is not a law I
passed, it is to signal that it is a serious part of what it
means to be part of this country. Here was a bipartisan passed
Senate resolution that said, well, no, I guess we don't need it
after all. It is not to be mandatory.
So when I said publicly I was disappointed in national
political leaders, that was not a partisan statement. This was
a very bipartisan statement. So I would have to ask you if you
want to say that I have politicized the census, I need to hear
from you the exact quote, either in a press conference, before
a hearing or in any other public setting where I have blamed
any Republican leader, and I don't believe that you will find
that quote.
Mr. Miller. We have several quotes that we will give to
you. I don't think maybe you used the word Republican, but you
say, ``Here is a moment when our national leadership could have
explained.'' The inference is to the Republicans, and when the
articles come out in the paper, they come out different than
maybe you think that they come out. ``A garbled message was
sent.'' ``Here is a moment when our national leadership could
have explained what serious role this information is in our
economy or society. That voice was either silent or it was
pandering to talk show hosts.'' That was before the Census
Advisory Committee. When you talk about the InterSurvey, the
inference was it was because of the remarks. The remarks were
on March 30, and the survey showed long form privacy concerns
jumped to 18 percent prior to March 30. So what happened was,
when I look at the data, when the forms got in the mail, people
received them, and then had concerns about privacy. It was
after the forms arrived, that is the 18 percent, and then
afterwards there were some comments by Senator Lott and
Governor Bush.
But I think you have been repeating--blaming in effect
Republicans for pandering to talk show hosts. None of us can
control talk show hosts. They get under my skin, too. But there
are articles in several papers. Here is one from the Fort
Lauderdale Sun Sentinel. ``Some Republican leaders view the
census as an invasion of privacy and urge Americans not to
answer questions that they consider too personal. That pulled
the entire response rate down for the country, Prewitt said
Wednesday.''
Mr. Prewitt. That is an incorrect quote. I did not say
that.
Mr. Miller. That is what is being reported. I think you
have already said that if you don't want to fill out a
question, at least give us enough information for apportionment
purposes. I used the illustration that my neighbor doesn't want
to give her phone number and income--fill out the rest. You
know I have been advocating for people to complete the forms as
best they can, and I know you never accused me of anything--but
let me switch to some questions now.
Yesterday the whole world seemed to come under attack from
a major computer virus which paralyzed computers. Were census
2000 operations affected in any way? I got some on my e-mail,
``I love you'' stuff. It was on the national news. I am just
curious if it had any impact on the Census Bureau.
Mr. Prewitt. No. We did a lot of work on our computers in
headquarters. Somehow we put down a message through all of our
computers, an anti-virus protection, and there may have been
isolated instances where isolated computers had read that
message, but there is certainly nothing of a large scale to
report at all.
Mr. Miller. It has been reported that organizations around
the country have had some real problems.
As you know, we have discussed the Tampa office before and
you responded in a letter to me this week. I visited my local
office in Bradenton, and Mrs. Maloney talked about an article
in Boston, there was an article in my local Bradenton newspaper
talking about a census worker working on the census in 1950,
and it was an interesting human interest story. I had one woman
who worked on the 1940 census and it was different back then.
They didn't use the mail response certainly in 1940. When did
mail come in, 1960?
Mr. Prewitt. 1960 was the first partial mail back.
Mr. Miller. But in 1950, they were knocking door to door
for everyone. It is more anecdotal, but I think my local office
in Bradenton is doing a good job. They have some difficult
areas to count, too. In Tampa apparently they are having
problems. Do you rate local census offices? Is there some type
of rating scale to identify those problem ones, an A, B, C, D,
F type of scale? I don't want to say that Tampa is in that low
category, but if in a local area you have a problem that is
real?
How many local census offices would you consider being
problem offices in however you want to define a census problem
office?
Mr. Prewitt. Fair enough. And I should say quickly what is
a problem local office varies from operation to operation.
Indeed, in the Bradenton-Tampa area, the mail back response
rate was quite strong and yet in other areas we had lower rates
than we had hoped. We had an LCO which wasn't doing as well as
we hoped in update/leave. So it is not like a single office
through all operations is particularly weak. In the Tampa
office, as I have written to you, we believe we had a serious
management problem. When you have a serious management problem
exactly at the recruitment period, that accumulates.
I would say across the country well under 5 percent had the
combination of those two things a management problem plus a
recruitment problem. The only thing you could do at that point
is try to change the management quickly. When we changed the
Tampa management, our rate shot up. We feel very good about the
quality of the staff. I think the press coverage in Tampa has
been reasonably consistently negative. We believe that we know
why that is so. We do not think it is about our operations, it
is about some other things. We remain disappointed that the
person who has gone to the press so often, who was an employee,
and then had to be let go, has not signed the release so we
can't explain why he was let go and that puts us at a
disadvantage in this press battle.
To your more general question, we look at these data of
course every day, our recruitment data, and right now we have
about 16 offices, that is as of a day and a half ago, 16
offices which we are particularly concentrating on with respect
to our recruitment system.
That 16 by tomorrow could be down to 8 because what happens
in some of these cases is that your payroll system is catching
up with you. Our data base is primarily our payroll system. We
have two offices where we had the very happy occurrence of a
large number of people shifted from update/leave and other
kinds of operations to enumerators. We were still paying them
on the old payroll, so it looked like we didn't have anyone
there, but we were fully staffed. It took 2 or 3 days to move
those records onto our NRFU payroll system. I would say that
the total number of offices right now about which we have any
serious concern are in the handful.
Now, tomorrow it may be a different set because we may have
a higher attrition rate than we expected. At any time the
probability of there being somewhere in the neighborhood of 10
to 15 offices is high.
Mr. Miller. Is recruitment the main way you tell?
Mr. Prewitt. How many people showed up at the training,
etc. Next week it will be attrition rates. If we have higher
than expected attrition rates, that will be the thing then.
Mr. Miller. You made this comment in your opening statement
but I think it would be nice if you elaborated. You are hiring
more people than you need, and some people are not going to get
called even though they may be qualified people. With an
operation of this size, communications is not always as ideal
as you would like it to be. So it is not always possible to let
people know why they are not getting called, could you just
elaborate on that?
Mr. Prewitt. Certainly. It has been an issue throughout
this entire process.
Mr. Miller. And Members of Congress are going to get these
calls at their offices, too.
Mr. Prewitt. Going back to your opening comment about
expecting us to use every tool in our tool box to make sure
that we have the highest level of accuracy possible, means for
us, we do not want to take any chance of diminishing the
recruitment pool until we are certain we don't need someone.
The recruitment pool has to be targeted at bilingual people and
people with a cultural understanding and people who understand
complicated situations in different parts of the country. We
have to find the right number of people and we are talking
about mail back response rate, and we may be at an LCO where we
are going to have to use all six callbacks and others where we
may get people more quickly because it has a high retirement
rate, etc.
From our point of view, the most important thing is to
retain that recruitment pool until we know we don't need it. We
are not calling people and saying we don't think that we are
going to need you. Even after nonresponse followup, we have a
very large operation called coverage improvement followup, we
need a very large field staff to do that task. We are not sure
where that task is going to fall. That is our national
estimate, but that will be concentrated in certain areas. We
want a recruitment pool there.
So all we can do--we would rather suffer the burden of some
people who are disappointed that they were not hired than not
have enough people to finish the census and that is simply the
position we have to take.
Mr. Miller. Thank you.
Mrs. Maloney.
Mrs. Maloney. Thank you, Mr. Chairman.
Mr. Chairman, in your opening statement you made reference
to a fax which Representative Coburn provided to the press, an
illegal act if the Privacy Act applied to Members of Congress,
I might add.
Dr. Prewitt, to try and put this incident in perspective,
the information inadvertently faxed to the wrong number was
not, as I understand it, title 13 material, was it?
Mr. Prewitt. That is correct. We fax no title 13 material.
Mrs. Maloney. What exactly is title 13 material?
Mr. Prewitt. Material which has a census response,
including an address. All of that material is simply handled
differently.
Mrs. Maloney. How is it handled differently?
Mr. Prewitt. It is only handled by people who are sworn
fully. No one can have any access to any sort of confidential
title 13 material that has not been sworn as a census employee.
When the actual forms come in, they are recorded in our
local office by sworn people. They are boxed, put into the
highly secure Fed Ex system and they come to our data capture
centers, and they are opened by sworn employees in our data
capture centers.
Mrs. Maloney. Do you have any idea how many faxes the
Bureau sends out in 1 day from its 520 local offices, 12
regional offices, 4 data capture centers and headquarter
offices by the 500,000 people currently on the payroll? Do you
have any sense of the proportion?
Mr. Prewitt. It is a very large number. A very large
number.
We regret any human error. Human error does occur. In this
particular instance the woman who made the call immediately
recognized that she had misdialed and tried to immediately
track the misdial. When we actually were able to reach the
woman, we asked that this material be destroyed immediately,
and that was refused by the woman who received it. Instead she
chose to share it. We then called the Congressman's office,
asked him to destroy the material immediately, and he also
suggested that he was not going to do that.
We are regretful that this piece of information got out.
Look, I am not trying to defend human error but I am very
pleased insofar as errors have occurred--and they will continue
to occur--thus far there has been no title 13 information which
has at all moved into any kind of public setting.
Mrs. Maloney. The Bureau has now had some limited
experience with the nonresponse followup. Do you have any
reports of hostility, of slammed doors and is any--what is the
response like? Is it more hostile than 1990? Have you had any
sense of a comparison or is it more friendly? What is the
response?
Mr. Prewitt. We are very pleased with the successful launch
of nonresponse followup. That is the training programs all
occurred on schedule and were fully staffed. Everyone--the
number of people that we needed came to our training sessions.
And we are now in the field. We only have 3 days of
information, of course, but approximately 8 percent of our non-
response followup workload is already completed in the field.
Now, that still has to be checked in and so forth. But from the
field point of view, they've now finished slightly over 8
percent of the cases. That's as of last night. We're right on
schedule with respect to that. We're certainly getting reports
of concerns, slammed doors and so forth. It's very anecdotal.
I have no way of knowing whether it's larger or smaller
than we got in 1990. The little factoid I learned yesterday is
we've had 212 dog bites so far, and one sort of serious bee
sting. But I don't have the base of that for 1990, whether
that's a higher rate of dog bites than 1990 or not. But we
worry about those kinds of things.
We do know in Anchorage, at least I read in the Anchorage
press, insofar as we can trust the press on these kinds of
things, at least four different enumerators in our update/leave
operation were met by people carrying guns and asked them not
to come on the property, so they left. But again, that's
anecdotal. I don't have a 1990 base to know whether this is
higher or lower than 1990.
Mrs. Maloney. Of the 41 million households in the non-
responsive followup of the universe, how many of them are long
forms and how many of them are short, do you know?
Mr. Prewitt. I would have to do the arithmetic quickly. It
should have been, of course, one out of six exactly, but since
the long form differential is 10 percent, if somebody could
quickly do that arithmetic for me. The point is--obviously the
point is that there are a higher percentage of long form cases
than we had anticipated.
Mrs. Maloney. The same proportion. You stated in your
testimony on page 2 that you're concerned about potential loss
of data due to opposition to the long form, and you stated
further in your testimony that you have no information on item
by item non-response, but do you have a sense of which
questions would cause the most problem if they weren't
answered?
Mr. Prewitt. Well, the most important information we have,
Congresswoman Maloney, is the 1990 item non-response pattern.
We think that's reasonably predictive of what we might get in
2000, and item non-response in 1990 varied from as little as
1\1/2\ percent to--on the income question, I believe the key
income question was 14 percent, but I don't have that in front
of me. I don't want to give you the wrong number. So it will
vary a lot.
We think that should be the most predictive. As I've said
in Congressman Rogers' hearing, that's what we'll be examining.
I don't--for this kind of work I don't believe--I don't
disbelieve in survey data, but I don't want to rely on survey
data. If you actually look at the InterSurvey question, when
they asked the respondents which questions do they find to be
intrusive, they found a very high percentage of people saying,
I think, for example, 22 percent said that they thought the
race question was intrusive. On the other hand, in 1990, only 2
percent of the American public did not answer the race
question. So I simply don't think that the surveys are likely
to be predictive of item non-response. What is most predictive
is the 1990 pattern.
Mrs. Maloney. Regarding the difference in response rates
for the long and short forms from the 1990 census and the 1988
and the 1998 dress rehearsals, could you explain and expand on
what those response rates were?
Mr. Prewitt. Yes. In the 1990 dress rehearsal, the response
rate--the differential response rate across a couple of sites
averaged about 6 percent and the non-response--the differential
in 1990 was 6 percent. In 2000, the differential response rate
between the long and short form was quite a bit higher. It
varied between whether it was update/leave in Columbia, SC, and
so forth. But it's not inaccurate to say that it would have
been close to 12 percent, and of course, 12 percent is the non-
response--is the differential in the 2000 pattern thus far.
Mrs. Maloney. What's your analysis of the roughly 12-point
differential in the long and short form response rates, and
what impact did it have on your planning for the 2000 census?
Mr. Prewitt. I'm sorry, Congresswoman Maloney. Would you
repeat that.
Mrs. Maloney. There was a differential of roughly 12
percentage points between the long and short form response
rates in the 1998 dress rehearsal, and what impact did that
have, if any, on your planning for the 2000 census?
Mr. Prewitt. We did not treat the differential in the dress
rehearsal as predictive of what we would get in 2000. So we did
not focus on that differential as a likely clue as to what
would happen in the census environment. We simply--we used the
dress rehearsal, of course, to test operations, not to try to
predict the behavior of the entire American public because
these are only three sites.
Mrs. Maloney. It was basically an operational run-through.
Mr. Prewitt. And we changed some operations, including, of
course, the second mailing based upon our dress rehearsal
experience. I might say, if I could, the approximate non-
response followup workload was 33 million short-form and 9
million long-form respondents.
Mrs. Maloney. Thank you. Thank you very much.
Mr. Miller. On the long form, one thing I congratulate the
Bureau for doing--I think it was actually before both of us
arrived on the scene--specifically was seeking professional
expertise to help design the form, and in reflecting back on
the 1990 form compared to this, I commend the Bureau for
getting professional surveyor consultants in helping do that. I
think that's positive.
Let me ask a question about the long form. The Bureau is
using one out of six for the long form. What criteria was used
for that? Why were you using one out of six? What is the
purpose of that?
Mr. Prewitt. Well, the real question is at what level of
geography do you want to be able to provide reasonably reliable
estimates? By doing one out of six, we can take our statistical
estimates down to a population of less than 20,000. So a
community of less than 20,000 or any other kind of group of
less than 20,000, that is, how many disabled veterans there
are, if that population is as large as 20,000, we would be able
to give you, the country, a reliable estimate of its
characteristics. At a higher sample, if we did one out of two
across the country, we could drive that 20,000 down to 12,000
or--I better get my experts to tell me exactly where. But
that's the reason. We thought that was a prudent way to help
the country understand the social dynamics, the housing
characteristics, population characteristics, and so forth.
[The information referred to follows:]
[GRAPHIC] [TIFF OMITTED] T0439.064
Mr. Miller. This was, I guess, before both of us were
actively involved in this. Let me bring up the issue of
Representative Coburn. I know it was an accident, and everybody
regrets accidents, but my understanding, by the way--is that
the information was not given to the press. For title 13 data,
I'm glad we have those standards, but I guess there's a
different standard for privacy data, which is individual Social
Security numbers and things like that, for which you don't have
the same level of security concerns for.
Mr. Prewitt. We have a very high level of security
concerns, Mr. Chairman, for all of our data. We simply have
different ways of processing non-title 13 and title 13 data. We
do use--you cannot, as I think Congresswoman Maloney's
questions implied, you cannot manage a census without using e-
mail, without using faxes, without using various forms of
distributing information around to the different actors who
need it. And therefore, to say that we would never use the fax
system or an e-mail system or administrative records would
cripple the census enormously.
So we do handle certain kinds of things differently from
how we handle title 13 data. We have an enormously high
standard for how we handle title 13 data. That doesn't mean we
don't have privacy concerns and security concerns for other
privacy data. Indeed, I'm sure that's true in the U.S. Congress
as well. But at a certain point, you do have to use the
apparatus that's available in the society for communication,
and faxes happen to be one of them, and faxes are subjected to
the human error of misdialing a number.
Mr. Miller. Let me ask a question about quality control
issues. We had one computer error where the prenotification
letters had the extra digit. We had the surname issue problem
for certain residential area units. We have confidence that the
quality control efforts are doing the right job, and especially
as we go into this non-response followup. Would you discuss
quality control issues and specifically quality control for the
enumerators? How do we know, for example, that an enumerator
who is assigned to go out and call on these 20 houses doesn't
go home and fill out 20 forms and bring them back to you?
I think we need to have assurances as there has been in the
past that there are quality control checks, and because we've
had some quality control failures, we are going to hopefully
avoid these in the future.
Mr. Prewitt. These are quite separate kinds of quality
control procedures of course, quality control on our software.
I would like to put in context the digit error that has been
discussed so much, and we, of course, brought that to your
attention immediately. We have now produced operations that
rest on about 2,500 different software programs, and I can't
promise you that there won't be other errors, but I can tell
you that all of the operations to date using about 2,500
different software programs are now completed and on schedule,
on budget, and correctly.
And if, in that huge amount, we did have a digit problem
with respect to a contractor, it happened, we tried to explain
how that happened and so forth. And then the second one that I
brought to your attention with respect to the surname which has
a very, very tiny operational implication, but nevertheless I
wanted you to know about that, you do have to see that as 1 out
of 2,500, and the fact that all the rest of them have
functioned as we had hoped for them is, to us, a very good
sign.
Now, the second issue that you raised, the issue of quality
control assurances with respect to enumerator work, certainly
the Census Bureau has been preoccupied throughout its history
with fabricated responses by enumerators. So we put in place
quality checks, and the work of every enumerator is double-
checked, that is, we either send someone back out or we use a
phone system to go back into the field and check on a
proportion of every enumerator's work on a regular basis. And
if we find any enumerators have reported to us a case, we go
back out and find out that that was a fraudulently provided
case. All of that enumerator's work is redone, all of it and,
of course, that enumerator is fired immediately. If you want
the actual rate at which we do that checking, Marvin Raines can
explain that better than I can. Would you like to hear that?
Mr. Miller. Yes.
Mr. Prewitt. It's 5 percent of the workload of every
enumerator. How frequently are we doing that on a consistant
basis? Every workload that comes in from an enumerator, 5
percent is pulled out as a sample and we go back and do a
quality check. So that is happening every day.
Mr. Miller. Let me ask a question about Social Security
numbers and clarify what the Bureau's position is because we
also want to caution people that there are going to be people
out there that are going to fake being census takers. But one
of the questions you are not asking----
Mr. Prewitt. I appreciate the opportunity to clarify that,
Mr. Chairman, because there are scam artists out there who are
trying to get Social Security numbers, bank card numbers, all
kinds of numbers calling themselves census employees.
Mr. Miller. Once again, by the way, describe what
identification a Census Bureau employee would have, so when
they are out there, they know they are not getting a scam
artist.
Mr. Prewitt. Let's do that first and then talk about Social
Security. Every enumerator, of course, has a badge. Every
enumerator is also carrying what we call a tote bag which has
the logo on it, and here is the badge. And every enumerator
also has his or her address file book, which is an 8\1/2\ by
11, 14--it's bigger than that. Sorry. It's not the kind of
thing that would be easy to fabricate and it has their work
materials.
Most importantly, every enumerator is expected to have
immediately available the phone number of the local office, so
a respondent can say, when you knock on the door, you say
you're from the Census Bureau. I don't know if you are from the
Census Bureau. You say, look, here's the phone number. Go call
the local office. Here is my name, here is my ID. And you can
double-check. Then you can go and check.
Most importantly, no enumerator should ever ask to come
into the home. Most people who are scamming, especially people
who are trying to conduct an act of thievery, need to get into
the home. And therefore if anyone asks to come into the home,
we're telling the American public that is an alert to you that
that is not a census taker. That doesn't mean you can't invite
them in.
Of course, enumerators get invited in and get served tea
and cookies. That's all very nice. Sometimes it doesn't happen
that way, but it does happen on some occasions. But
nevertheless, no one should ever ask to go into the home.
That's extremely important. Now, there will, nevertheless, be
scam artists out there trying to get information from a
household of a sort that could be used against them.
With respect to the Social Security issue during the
mailout phase approximately 21,000 households got a special
letter from me--four different versions of that letter, saying
that this is the census, and for various complicated reasons,
we're going to be asking your Social Security number, and there
are four different treatments in that 21,000, depending upon
the experimental design. And we made it quite clear this was
voluntary. This was not mandatory.
This was not part of the usual decennial census procedures
itself, but we were asking that question as a test for a
limited number of households. The reason we did that experiment
in the context of the census environment is because we were
under strong injunction from the U.S. Congress, and indeed, you
referenced it again in your opening comments, to investigate to
what extent we could use administrative records more
efficiently than we're doing in 2000. The Census Monitoring
Board had a full hearing on administrative records. Part of the
administrative record system of this country, of course, is
Social Security numbers.
So we were doing that as a way to test the privacy
concerns, and we'll report, of course, our evaluation of that
experiment as soon as that's been completed. That won't be
until sometime next year. So in those cases, we actually ask in
the census environment for a Social Security number for roughly
21,000 households making reference to the fact that, in
addition, we actually ask for the Social Security number in our
Survey of Income and Program Participation, our SIPP survey,
and that's in order to actually strengthen the survey
instrument, and because we are under title 13, we are allowed
to cooperate with other agencies and strengthen the data base
by sharing reports.
Mr. Miller. What's the sample size of that?
Mr. Prewitt. Sample size of SIPP is 36,000 households.
Mr. Miller. Correct me if I'm wrong. None of the non-
response enumerators will ask Social Security numbers?
Mr. Prewitt. That's the key part of your question. During
non-response followup, no enumerator has any reason ever to ask
for a Social Security number, because the experimental work we
did was only in terms of mail-back response rates. It was never
intended to be part of nonresponse followups. You're correct.
No enumerator has any reason to ever ask for a Social Security
number of anyone in the society.
Mr. Miller. Let me ask one final question. It's hard to
enumerate areas. Does each local census office have a written
plan for dealing with the hard-to-count neighborhoods? Everyone
is different. You were talking about Mrs. Maloney's district is
one of the hardest to count. My hard-to-count areas are the
migrant areas more in the center part of the State, actually
even outside of my congressional district. Do local offices
have specific plans to address their specific problems?
Mr. Prewitt. You're quite right that a hard-to-count gated
community can be just as hard to count as a migrant worker
community. And, yes, sir, every LCO does have its hard-to-count
strategy. This is part of the record because we put this
material as an appendix into my written testimony, and it does
indeed take into account those kinds of things, languages
spoken, distance the enumerator has to travel, is it very
remote, things like gated communities.
Mr. Miller. Each office would have a little different plan.
Mr. Prewitt. Exactly. There's a whole list of the traits,
but they weigh very differently office to office. It's not a
cookie cutter operation.
Mr. Miller. With regards to oversight, we can have access
to it when we visit a local office to see what----
Mr. Prewitt. Yes, sir.
Mr. Miller. One comment on the hard-to-count. Would you
comment, on what's happening on Indian reservations in
particular?
Mr. Prewitt. Let me start, if I can, with remote Alaska
because the number is clearest in my mind because I just talked
to the people up there who completed that. We are now completed
with remote Alaska, and every village in which the local
leadership, which is a vast majority of them, cooperated with
the census. We completed 100 percent of the count. We're very
pleased with that work thus far. That's a part of our American
Indian and Native Alaskan populations.
I think with respect to Indian land more generally,
overall, the pattern has been very strong, very positive. There
are two or three pockets, and I will have to ask Marvin Raines
to comment in detail. Two or three pockets where we are still
getting some resistance. I think there is one in Montana, as I
recall. This is not a general problem. Indeed, the mail-back
response rate from some of the Indian areas beat their ``plus
5'' goal. About as many of those as did across the country. 17
percent of communities across the country met the ``plus 5''
goals. It's an extraordinary accomplishment by those
communities.
Does anyone know offhand the proportion of those who are in
areas?
Mr. Miller. Let us just get that information.
Mr. Prewitt. We'll give it to you.
Mr. Miller. The American Indians were one of the most
undercounted populations we had in the 1990 census.
Mrs. Maloney.
Mrs. Maloney. Thank you. On administrative records, Dr.
Prewitt, can you use administrative records without Social
Security numbers, or do you need Social Security numbers?
Mr. Prewitt. In principle, there certainly are Social
Security numbers--excuse me--there are administrative records;
for example, school attendance records, perhaps occupancy
records from the local government, which would not necessarily
require you to use a Social Security number. That would be very
uneven across the country.
When we looked at administrative records, one of the things
that we found was it is very difficult to implement anything
that was standard across the country because different
jurisdictions do not keep the same kind of records. Our school
attendance records, our housing occupancy records, our housing
start records, all kinds of other records are different from
jurisdiction to jurisdiction, so it's very difficult to design
a census in a way that standardizes quality across the United
States. The only things which are standardized across the
United States are largely Federal programs such as Medicare,
Medicaid and those programs all do use--I think all of them use
Social Security numbers as part of their data record.
I might say, if I could say another word or two on this,
Mr. Chairman, you asked what was the Census Bureau's position
on Social Security numbers. We have no position. Indeed given
the concerns about privacy in this country, we have never
recommended, and I don't think we would ever recommend, that
this country have a national identification number system. The
census is done in Scandinavian countries, for example, based on
a national identification number system. My own judgment would
be that that would not be a direction that either the U.S.
Congress or the Census Bureau should move toward.
Now, there's a very complicated issue, because if we don't
have a national identification system and yet we're under
pressure to use administrative records in order to keep costs
down and improve coverage, what is the nature of the
administrative records that we can use which stop short of what
the American public could interpret as a national
identification number, which is to say, a Social Security
number? So it's a very tough question that the Congress will
have to discuss as we start planning for 2010.
We did think we had an obligation to the Congress to sort
of try to learn what we could in the census environment. It's
very difficult to learn some of these things outside of the
census environment. That's why we conducted the experiment.
It's not a policy position of the Bureau to recommend that we
use administrative records in the way that would necessarily
incorporate Social Security numbers as part of it.
Mrs. Maloney. The chairman has repeatedly mentioned that he
would like to see administrative records used more but that
really basically raises a privacy concern because part of
administrative records, the reliable ones, Medicare, Medicaid
which you mentioned nationally, all involve a Social Security
number which is a privacy concern. So there is a privacy
concern directly related to administrative records. Is that
what you're saying?
Mr. Prewitt. Yes. Certainly at the national level there
would be.
Mrs. Maloney. I'm glad that Chairman Miller clarified that
Congressman Coburn did not give census information to the
press, but based on his strong statements on privacy, it would
be important, I think, to have the same privacy level for
Members of Congress, as other agencies, such as the Census
Bureau, and I think something that we could work on in a
bipartisan way is a bill that would cover Congress under the
Privacy Act and have that go through Congress so that Congress
people were held to the same privacy standard, because privacy
is very important. That could be something we could work on. I
would certainly support it.
All I can say, Dr. Prewitt, is congratulations. I'd like to
publicly thank you and all of the professionals and part-time
workers, full-time workers in the Census Bureau. You have
reversed three decades of decline, and I have no further
questions at this point. I just congratulate you and wish you
well during this difficult enumeration stage and just really
hope that everyone will cooperate with the enumerators and help
us get the most accurate count in America. Thank you very much.
Mr. Miller. In conclusion, let me say thank you. It's
satisfying at this stage because of the mail response, which,
as you know, is one of the most difficult parts of it. Things
are looking good. I'll be looking forward to progress reports
as we go through this process. We'll have little bumps along
the way, we all know. You're going to have an employee that's
not going to be one that's going to live up to the standards of
the Bureau, and that's going to be an embarrassment, but we
need to prepare for that too.
On behalf of the subcommittee, thank you for the job you're
doing and thank you for being here today.
I ask unanimous consent that all Members' and witnesses'
written opening statements be included in the record. Without
objection, so ordered. In case there are additional questions
that Members may have for our witnesses, I ask unanimous
consent that the record remain open for 2 weeks for Members to
submit questions for the record and that the witnesses submit
written answers as soon as practicable. Without objection. So
ordered. Meeting adjourned. Thank you.
[Whereupon, at 11:30 a.m., the subcommittee was adjourned.]