2000 Census: Answers to Hearing Questions on the Status of Key Operations
(Correspondence, 05/31/2000, GAO/GGD-00-109R).

Pursuant to a congressional request, GAO provided information on the
status of the Bureau of the Census' key census operations, focusing on:
(1) whether the Bureau followed GAO's recommendations and adopted an
alternate form of contingency planning instead of relying on Congress
for a supplemental appropriation; (2) why the census is such a local
endeavor; (3) whether the Bureau will be able to translate the high
level of public awareness into participation for the 2000 Census; (4)
whether partnership specialists will be stretched too thinly to have a
successful impact on the 2000 Census; (5) the challenges facing the
Bureau in conducting a timely and accurate followup; (6) how the Bureau
could intentionally or unintentionally cut corners to get the
nonresponse follow-up workload done in a shorter period of time; and (7)
the risks that could jeopardize the release of timely data.

GAO noted that: (1) in his written response to GAO's recommendation, the
Secretary of the Department of Commerce noted that Bureau officials
share GAO's "awareness of the challenges inherent in conducting a
complete and thorough nonresponse follow-up operation;" (2) however, the
Secretary added that the current plan for the 2000 Census will produce
the most accurate enumeration possible, and that the Bureau must devote
its full attention to carrying out each component of that plan; (3) the
Bureau noted that the only serious contingency would be to request a
supplemental appropriation; (4) the census is in many respects a local
endeavor because the key ingredients of a successful population count,
such as a complete and accurate address list and timely and accurate
data collection, are carried out by locally recruited census employees
going from one neighborhood to the next; (5) national-level data,
although useful for providing an overall perspective on the census,
tends to obscure operational challenges and successes at the local level
that can affect the quality of the census; (6) the response rate to the
2000 Census was 65 percent--4 percentage points higher than what the
Bureau had anticipated and equal to the 1990 Census response rate; (7)
still, preliminary data suggest that the Bureau was unable to translate
high levels of census awareness into census participation; (8) although
the response rate was 65 percent, various polls have suggested that the
public's awareness of the census was significantly higher; (9) Bureau
partnership specialists appear to be more thinly stretched, on average,
for the 2000 Census, than they were for the 1998 dress rehearsal; (10)
completing the nonresponse follow-up workload within the allotted
timeframe will be critical to collecting quality data because the census
is progressing on a very tight schedule and the Bureau needs time to
complete other census operations, including coverage evaluations that
will be used to estimate census undercounts and overcounts; (11) the
Bureau's nonresponse follow-up workload for 2000 is about 42 million
housing units, and it has scheduled ten weeks to complete the operation;
(12) in planning the 2000 nonresponse follow-up operation, the Bureau
took steps to avoid the problems it encountered in 1990; (13) to address
expected turnover, the Bureau planned to hire two people for each of its
146,000 enumerator positions; and (14) any material delay in checking-in
census forms, both long and short, could adversely affect the timeliness
of downstream activities, such as determining nonrespondents and
tabulating final results.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  GGD-00-109R
     TITLE:  2000 Census: Answers to Hearing Questions on the Status of
	     Key Operations
      DATE:  05/31/2000
   SUBJECT:  Census
	     Population statistics
	     Data collection
	     Public relations
	     Government information dissemination
	     Data integrity
	     Reporting requirements
	     Federal/state relations
IDENTIFIER:  2000 Decennial Census
	     1990 Decennial Census
	     Pomona (CA)

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************
GAO/GGD-00-109R

United States General Accounting Office
GAO

GAO/GGD-00-109R

Ordering Copies of GAO Reports
The first copy of each GAO report and testimony
is free. Additional copies are $2 each. Orders
should be sent to the following address,
accompanied by a check or money order made out
to the Superintendent of Documents, when
necessary. VISA and MasterCard credit cards are
accepted, also. Orders for 100 or more copies to
be mailed to a single address are discounted 25
percent.
Order by mail:
U.S. General Accounting Office
P.O. Box 37050
Washington, DC 20013
or visit:
Room 1100
700 4th St. NW (corner of 4th and G Sts. NW)
U.S. General Accounting Office
Washington, DC
Orders may also be placed by calling (202) 512-
6000 or by using fax number (202) 512-6061, or
TDD (202) 512-2537.
Each day, GAO issues a list of newly available
reports and testimony. To receive facsimile
copies of the daily list or any list from the
past 30 days, please call (202) 512-6000 using a
touch-tone phone. A recorded menu will provide
information on how to obtain these lists.
Viewing GAO Reports on the Internet
For information on how to access GAO reports on
the INTERNET, send e-mail message with "info" in
the body to:
[email protected]
or visit GAO's World Wide Web Home Page at:
http://www.gao.gov

Reporting Fraud, Waste, and Abuse in Federal
Programs
To contact GAO FraudNET use:
Web site:
http://www.gao.gov/fraudnet/fraudnet.htm
E-Mail: [email protected]
Telephone: 1-800-424-5454 (automated answering
system)

United States
General Accounting Office
Washington, D.C. 20548-0001
Official Business
Penalty for Private Use $300
Address Correction Requested

                    Bulk Rate
               Postage & Fees Paid
                       GAO
                 Permit No. G100

(410564)

United States General Accounting Office General Government
Division
Washington, D.C.  20548

Page 1GAO/GGD-00-109R Questions on the Status of Key Census Op
erations

B-285273

May 31, 2000

The Honorable Dan Miller
Chairman, Subcommittee on the Census
Committee on Government Reform
House of Representatives

Subject:  2000 Census:  Answers to Hearing Questions on the
Status of Key Operations

Dear Mr. Chairman:

This letter responds to your request for additional
information on census operational issues following the
Subcommittee on the Census' February 15 hearing on the status
of key census-taking activities.  The enclosure contains our
response to questions that we received from the Subcommittee.
Because our response is based primarily on our previous work,
we did not obtain comments from the Department of Commerce on
a draft of this letter.  However, we asked senior Bureau of
the Census officials to review the letter's technical
accuracy.  In their May 24, 2000, reply, Bureau officials
provided us with additional information on problems that
occurred during the 1990 nonresponse follow-up operation and
the steps the Bureau took to avoid similar difficulties for
the 2000 Census. The Bureau also gave us further perspective
on its use of proxy data.  We incorporated this information as
appropriate in our response to questions 6 and 7.

We are sending copies of this letter to the Honorable Carolyn
B. Maloney, Ranking Minority Member, Subcommittee on the
Census; the Honorable William M. Daley, Secretary of Commerce;
and the Honorable Kenneth Prewitt, Director of the Bureau of
the Census.  We will make copies available to others upon
request.  If you have any questions concerning this letter,
please contact me on (202) 512-8676.

Sincerely yours,

J. Christopher Mihm
Associate Director, Federal Management
 and Workforce Issues

Enclosure
 

Enclosure
Responses to Subcommittee Questions Following the
February 15, 2000, Hearing on Key Census-Taking
Operations
Page 6GAO/GGD-00-109R Questions on the Status of K
ey Census Operations
1.  The General Accounting Office (GAO) has
suggested that the Census Bureau adopt some
alternate form of contingency planning instead of
relying on the Congress for a supplemental
appropriation.  Have you received any
correspondence from the Census Bureau regarding
your requests and have you been asked to make any
recommendations?

In our December 1999 report, we recommended that
the Director, Bureau of the Census, develop a
contingency plan of actions the Bureau can take to
address the operational challenges that would
result from a questionnaire response rate that is
lower than anticipated.1  We also noted that the
Bureau's plan should (1) address the budgetary,
scheduling, staffing, and other logistical
implications of collecting data from a larger
number of nonresponding households and (2) include
options and procedures to balance the pressure to
complete nonresponse follow-up on schedule without
comprising the quality of census data.

The national, initial response rate to the 2000
Census was 65 percent-4 percentage points above
the 61-percent response rate that the Bureau had
anticipated.  However, not surprisingly, several
local census offices are facing lower-than-
expected response rates and, therefore, a larger
than expected follow-up workload.  Thus, at
certain local census offices, completing
nonresponse follow-up on schedule, without
compromising the quality of census data, could be
a concern.

In his written response to our recommendation, the
Secretary of the Department of Commerce noted that
Bureau officials share our "awareness of the
challenges inherent in conducting a complete and
thorough nonresponse follow-up operation."
However, the Secretary added that the current plan
for the 2000 Census will produce the most accurate
enumeration possible, and that the Bureau must
devote its full attention to carrying out each
component of that plan.  The Secretary's comments
echo those that the Bureau made on a draft of our
December report.  The Bureau noted that the only
serious contingency would be to request a
supplemental appropriation.

2.  Please explain why the census is such a local
endeavor, as I predominately hear reports from the
Census Bureau of how the census is progressing on
the national scale.

The census is in many respects a local endeavor
because the key ingredients of a successful
population count, such as a complete and accurate
address list and timely and accurate field data
collection, are carried out by locally recruited
census employees going from one neighborhood to
the next, often door-to-door.  Moreover, critical
tasks, such as building public awareness of the
census and motivating people to respond as well as
locating pockets of hard-to-count population
groups, are accomplished in large part by
partnerships between the Bureau and local
governments and community groups.  As a result,
national-level data, although useful for providing
an overall perspective on the census, tends to
obscure operational challenges and successes at
the local level that can affect the quality of the
census.

3.  In Mr. Mihm's testimony he talked about the
difference between public awareness of the census
and motivating the public to actually participate.
Based on data from the 1990 Census and the dismal
results from the 1998 dress rehearsal, do you have
any reason to believe that the Census Bureau will
be able to translate the high level of public
awareness into participation for Census 2000?  Do
you have any other recommendations?

As we previously noted, the response rate to the
2000 Census was 65 percent-4 percentage points
higher than what the Bureau had anticipated and
equal to the 1990 Census response rate.  The
Bureau's accomplishment in this regard is
particularly noteworthy given various attitudinal
and demographic trends, including concerns over
privacy and a larger non-English-speaking
population, that act against a high response rate.

Still, preliminary data suggest that the Bureau
was unable to translate high levels of census
awareness into census participation.  Indeed,
although the response rate was 65 percent, various
polls have suggested that the public's awareness
of the census was significantly higher.

We will continue our assessment of the Bureau's
outreach and promotion program and examine
possible refinements and recommendations as data
on the impact of the program become available.

4.  Mr. Mihm also stated that the Census 2000
"Partnership Specialists," the "Media
Coordinators," and the interaction of the Complete
Count Committees are key to the success of Census
2000.   Past experience from the 1998 dress
rehearsal tells us that the partnership
specialists were stretched too thin to make any
sort of difference.  Do you have any reason to
believe that things will work better in this
regard for Census 2000?

Bureau partnership specialists appear to be more
thinly stretched, on average, for the 2000 Census,
than they were for the dress rehearsal.2  The
Bureau hired over 600 partnership specialists to
initiate and sustain local outreach and promotion
initiatives, including Complete Count Committee
activities.  According to the Bureau, there are
around 12,000 Complete Count Committees; thus, on
average, each partnership specialist is
responsible for assisting approximately 20
committees.  During the South Carolina dress
rehearsal, the Bureau's two partnership
specialists were each responsible for assisting an
average of six Complete Count Committees.
However, some committees never formed, while
others became inactive, partly because the
Bureau's two partnership specialists were spread
too thin to provide meaningful assistance.

We are assessing the impact that partnership
specialists and Complete Count Committees had on
the census as part of our longer-term review of
the Bureau's outreach and promotion program, and
will report back to the Subcommittee as data are
available.

5.  Mr. Mihm testified that the Census Bureau
could be challenged to complete nonresponse follow-
up on schedule without compromising data quality.
Please explain this further.  What are the
immovable deadlines?

Nonresponse follow-up began as scheduled on April
27, 2000, and is to be completed 10 weeks later on
July 7, 2000.  Completing the nonresponse follow-
up workload within this time frame will be
critical to collecting quality data because the
census is progressing on a very tight schedule and
the Bureau needs time to complete other census
operations, including coverage evaluations that
will be used to estimate census undercounts and
overcounts.

Moreover, the Bureau has found that the quality of
data collected during its field follow-up efforts
declines over time in part because people move or
tend to forget who was residing at their
households on Census Day.  Further, to complete
nonresponse follow-up on schedule, the Bureau may
need to rely more heavily on data collected from
secondhand sources, such as neighbors.  Such data
are not as reliable as data collected directly
from household members.

6.  As I understand it, the Census Bureau has to
follow-up on roughly 46 million nonresponding
households during the nonresponse follow-up stage
of Census 2000.  They plan to do all of this in 10
weeks.  On the other hand, it took the Census
Bureau 14 weeks to finish nonresponse follow-up
for 34 million households in 1990. That seems like
an unreasonable time frame to do considerably more
work.  Please discuss how the Census Bureau could
intentionally or unintentionally cut corners to
get this larger workload done in a shorter period
of time.

During the 1990 Census, the nonresponse follow-up
operation was scheduled to last for 6 weeks-from
April 26 through June 6.  However, because of an
unexpectedly sharp decline in the mail response
rate, many local census offices had to cope with
follow-up workloads greater than those offices
originally planned to handle.  High turnover and
other staffing difficulties also hampered the
timely completion of nonresponse follow-up in
1990.  Thus, according to the Bureau, by June 4,
1990, the Bureau had finished just 70 percent of
its 34 million housing unit follow-up workload,
and did not fully complete the operation until
July 30, about 14 weeks after it began.

The Bureau's nonresponse follow-up workload for
2000 is about 42 million housing units, and it has
scheduled 10 weeks to complete the operation.
Thus, compared to 1990, the Bureau has less time
to complete a larger workload.  However, in
planning the 2000 nonresponse follow-up operation,
the Bureau took steps to avoid the problems it
encountered in 1990.  For example, to address
expected turnover, the Bureau planned to hire two
people for each of its 146,000 enumerator
positions.  Most local census offices met this
goal.

In addition, the Bureau's nonresponse follow-up
operation was helped by a higher-than-expected
response rate.  Indeed, because the response rate
was 65 percent as opposed to the anticipated 61
percent, the Bureau needs to collect data from
about 4 million fewer households than initially
planned.

Nevertheless, with a follow-up workload of 42
million households, the Bureau still faces a
tremendous task.  Therefore, as the 10-week time
frame allotted for nonresponse follow-up winds
down, it will be important for the Bureau to
monitor proxy data use to ensure that enumerators
are complying with proper procedures so as not to
reduce data quality.

7.  We have recently learned that there were
alarmingly high levels of proxy data collected
during the nonresponse follow-up stage of the
dress rehearsals for all three of the sites.   Why
did the Census Bureau have to cut corners and
collect proxy data to get out of the field so
early to finish nonresponse follow-up?

During the dress rehearsal, nonresponse follow-up
operations were completed on schedule in both
Menominee and Sacramento and 6 days ahead of
schedule in South Carolina.  However, the Bureau
relied more heavily on proxy data than it had
planned.  As shown in table 1, although the Bureau
hoped to limit the portion of the nonresponse
follow-up universe enumerated by proxy to less
than 6 percent, the Bureau did not achieve this
objective at any of the three dress rehearsal
sites.

Table 1: Dress Rehearsal Households Enumerated by
Proxy
Dress rehearsal site             Percentage of the
                                          occupied
                             nonresponse follow-up
                                          universe
                               enumerated by proxy
Sacramento                                   20.1%
South Carolina                                16.4
Menominee                                     11.5
Source:  U.S. Bureau of the Census.

According to the Bureau, one reason for the
comparatively high use of proxy data was that
obtaining interviews with household members proved
to be more difficult than the Bureau had
anticipated.

The Bureau has not set a specific goal for the
level of proxy data for the 2000 Census because,
according to the Bureau, data from the dress
rehearsal and other census experience were
insufficient to determine what a reasonable proxy
rate should be.  However, the Bureau wants to
minimize the use of proxy data as much as
possible.

8.  Mr. Mihm stated in his testimony that it may
cost at least $57 to enumerate each housing unit
for 2000 compared to about $31 in 1990.  This is
an increase of 84 percent in 1999 dollars.  Do you
agree that the Census Bureau could have planned
better to help mitigate this tremendous cost
increase?

We have not reviewed the extent to which better
planning could have mitigated the substantial
increase in census costs that the Bureau
experienced.  However, this issue will be examined
as part of our ongoing review of the 2000 Census
and lessons learned for 2010.

9.  Mr. Mihm testified during a recent data
processing test in Pomona, CA, Census 2000
employees were only able to check-in 54 percent of
their goal for census short forms.  Short form
questionnaires go to 5 out of 6 households across
the country.  If this situation is not remedied do
you foresee any significant risks that could
jeopardize the release of timely data?

Checking-in census forms, both short and long, is
one of a series of interrelated steps in the
Bureau's process for capturing census data; thus,
any material delay in checking-in forms could
adversely affect the timeliness of downstream
activities, such as determining nonrespondents and
tabulating final results.  To correct the problems
that surfaced during the Pomona site operational
test, the Bureau's contractor provided additional
training and practice time for the personnel who
perform the check-in activities.  Subsequently,
during the actual census, the operations at the
Pomona data capture center, as well as at the
Bureau's three other data capture centers, have
received and checked-in the required number of
forms.  For example, according to Bureau data, as
of May 14, the Pomona data capture center received
and checked-in over 24 million questionnaires,
which is about 2 million more than planned.

_______________________________
1 2000 Census:  Contingency Planning Needed to
Address Risks That Pose a Threat to a Successful
Census (GAO/GGD-00-6, Dec. 14, 1999).
2 The dress rehearsal for the 2000 Census was held
at three sites:  Sacramento, CA; 11 South Carolina
county governments and the city of Columbia; and
Menominee County, WI, including the Menominee
American Indian Reservation.  The dress rehearsal
tested the Bureau's operations and procedures
planned for the 2000 Census and was conducted in
April 1998.
*** End of document ***