Census Bureau Needs to Assess Data Quality Concerns Stemming from Recent Design Changes (GAO-21-142)
Summary
A GAO report to congressional addressees, GAO-21-142, dated December 2020, on the 2020 Census. It describes how the U.S. Census Bureau responded to COVID-19 in multiple phases, first suspending field operations in March 2020 for two successive 2-week periods. In April 2020 the Bureau extended the suspension to a total of 3 months for Non-response Follow-up (NRFU), the most labor-intensive decennial field operation, in which enumerators go door-to-door to households that have not responded. Its title states that the Bureau needs to assess data quality concerns stemming from recent design changes.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Full text
United States Government Accountability Office
Report to Congressional Addressees
2020 CENSUS
December 2020
Census Bureau
Needs to Assess Data
Quality Concerns
Stemming from
Recent Design
Changes
GAO-21-142
December 2020
2020 CENSUS
Census Bureau Needs to Assess Data Quality
Concerns Stemming from Recent Design Changes
Highlights of GAO-21-142, a report to
congressional addressees
Why GAO Did This Study What GAO Found
As the Bureau was mailing out The U.S. Census Bureau (Bureau) responded to COVID-19 in multiple phases.
invitations to respond to the decennial The Bureau first suspended field operations in March 2020 for two successive 2-
census and was preparing for fieldwork week periods to promote the safety of its workforce and the public. In April 2020,
to count nonresponding households, the Bureau extended this suspension to a total of 3 months for Non-response
much of the nation began closing down Follow-up (NRFU), the most labor-intensive decennial field operation that
to contain the COVID-19 pandemic. involves hundreds of thousands of enumerators going door-to-door to collect
In response to the pandemic, the census data from households that have not yet responded to the census. At that
Bureau has made a series of changes time, the Department of Commerce also requested from Congress a 120-day
to the design of the census. extension to statutory deadlines providing census data for congressional
Understanding the chronology of apportionment and redistricting purposes, and the Bureau developed and
events and the Bureau’s decisions, implemented plans to deliver the population counts by those requested
along with the factors and information deadlines.
sources that it considered, can help to
shed light on the implications and The Bureau implemented NRFU in multiple waves between July 16 and August
tradeoffs of the Bureau’s response. 9, 2020, to ensure that operational systems and procedures were ready for
nationwide use. The Bureau considered COVID-19 case trends, the availability of
This report, the first in a series of personal protective equipment, and the availability of staff in deciding which
retrospective reviews on the 2020 areas to start NRFU first.
Census, examines the key changes
that the Bureau made in response to On August 3, 2020, the Bureau announced that, as directed by the Secretary of
the COVID-19 outbreak and how those Commerce, it would accelerate its operational timeframes to deliver population
changes affect the cost and quality of counts by the original statutory deadlines. The U.S. District Court for the Northern
the census. GAO performed its work District of California in September 2020 issued an injunction that reversed the
under the authority of the Comptroller Secretary’s August 2020 directions for design changes and the Bureau’s
General to conduct evaluations on the adherence to the statutory deadlines, but the Supreme Court ultimately stayed
2020 Census to assist Congress with this injunction in October 2020 and allowed the Bureau to proceed with its August
its oversight responsibilities. GAO 2020 design changes. As a result, the Bureau shortened NRFU by over 2 weeks
reviewed Bureau decision memos, and reduced the time allotted for response processing after NRFU from 153 days
interviewed Bureau officials, and to 77 days.
consulted contemporaneous COVID-
19 case data for context on the GAO has previously noted that late design changes create increased risk for a
Bureau’s COVID-19 response. quality census. The Bureau is examining ways to share quality indicators of the
census in the near term and has a series of planned operational assessments,
What GAO Recommends coverage measurement exercises, and data quality teams that are positioned to
GAO is recommending that the Bureau retrospectively study the effects of design changes made in the response to
update and implement its assessments COVID-19 on census data quality. The Bureau is still in the process of updating
to address data quality concerns its plans for these efforts to examine the range of operational modifications made
identified in this report, as well as any in response to COVID-19, including the August 2020 and later changes.
operational benefits. In its comments,
the Department of Commerce agreed As part of the Bureau’s assessments, it will be important to address a number of
with GAO’s findings and concerns GAO identified about how late changes to the census design could
recommendation. The Bureau also affect data quality. These concerns range from how the altered time frames have
provided technical comments, which affected population counts during field data collection to what effects, if any,
GAO incorporated as appropriate. compressed and streamlined post-data collection processing of census data may
have on the Bureau’s ability to detect and fully address processing or other errors
before releasing the apportionment and redistricting tabulations. Addressing
View GAO-21-142. For more information, these concerns as part of the overall 2020 assessment will help the Bureau
contact J. Christopher Mihm at 202-512-6806
or by email at mihmj@gao.gov or Nick
ensure public confidence in the 2020 Census and inform future census planning
Marinos at 202-512-9342 or by email at efforts.
marinosn@gao.gov
United States Government Accountability Office
Contents
Letter 1
Background 2
Bureau Needs to Examine Effects on Data Quality of Its Design
Changes in Response to COVID-19 5
Conclusions 21
Recommendations for Executive Action 22
Agency Comments and Our Evaluation 22
Appendix I Comments from the Department of Commerce 25
Appendix II GAO Contact and Staff Acknowledgments 26
Tables
Table 1: Several 2020 Census Operations Relied on In-Person
Contacts by Design 3
Table 2: Examples of Changes to 2020 Census Information
Technology (IT) Contracts Due to COVID-19 20
Figure
Figure 1: The Census Bureau Adjusted Its Operational Timeline
during the COVID-19 Pandemic 6
Page i GAO-21-142 2020 Census
Abbreviations
ACO area census office
Bureau U.S. Census Bureau
COVID-19 Coronavirus Disease 2019
IT information technology
MAF Master Address File
NRFU Non-Response Follow-Up
PES post-enumeration survey
SBE Service-Based Enumeration
This is a work of the U.S. government and is not subject to copyright protection in the
United States. The published product may be reproduced and distributed in its entirety
without further permission from GAO. However, because this work may contain
copyrighted images or other material, permission from the copyright holder may be
necessary if you wish to reproduce this material separately.
Page ii GAO-21-142 2020 Census
Letter
441 G St. N.W.
Washington, DC 20548
December 3, 2020
Congressional Addressees
Last March, as the U.S. Census Bureau (Bureau) was mailing out
invitations to respond to the decennial census and was preparing for
fieldwork to count nonresponding households, much of the nation began
closing down to contain the Coronavirus Disease 2019 (COVID-19)
pandemic. 1
In response to the pandemic, the Bureau made a series of late changes
to the design of the census. These changes affected the way the Bureau
did its work as well as the time it took to do the work. The changes
introduced risks to the quality of data that the Bureau provides for
Congressional apportionment and redistricting purposes. 2 These data are
also central to allocating hundreds of billions of federal dollars.
Understanding the chronology of events that took place and decisions
that the Bureau made in 2020, along with the factors and information
sources that the Bureau considered, can help to shed light on the cost
and quality implications and tradeoffs of the Bureau’s response.
This report, the first in a series of retrospective reviews on the 2020
Census with the overall aim of informing planning and decisions for the
design of the 2030 Census, examines the key changes that the Bureau
made in response to the COVID-19 outbreak and how those changes
affect the cost and quality of the census. We performed our work under
the authority of the Comptroller General to conduct evaluations on the
2020 Census to assist Congress with its oversight responsibilities.
To address our objective, we reviewed Bureau decision memos and
planning documents. We also reviewed Bureau announcements related
to the COVID-19 response and any characterizations of the resulting
operational changes. We also interviewed Bureau officials to better
understand downstream effects of operational changes made during the
1The first mailed invitations to respond to the census were delivered on March 12, 2020,
the day after the World Health Organization declared COVID-19 to be a pandemic, and
the day before the U.S. declared COVID-19 to be a national emergency.
2Census data are used, among other purposes, to apportion the seats of the U.S. House
of Representatives and redraw congressional districts in each state.
Page 1 GAO-21-142 2020 Census
COVID-19 response, as well as to learn about any actions the Bureau is
taking to monitor cost and quality effects of the changes.
To place the key points of the Bureau’s COVID-19 response in
chronological context, we examined COVID-19 case data from the
Centers for Disease Control and Prevention. We found these data
sufficiently reliable for the purposes of our reporting objective. We also
accessed available data from the National Governors Association on
contemporaneous state-by-state policy actions taken in response to the
pandemic.
We conducted this performance audit from August 2020 to December
2020 in accordance with generally accepted government auditing
standards. Those standards require that we plan and perform the audit to
obtain sufficient, appropriate evidence to provide a reasonable basis for
our findings and conclusions based on our audit objectives. We believe
that the evidence obtained provides a reasonable basis for our findings
and conclusions based on our audit objectives.
Background
Pre-Pandemic Plan for For the 2020 Census, the Bureau is required by law to count the
Census Called for population as of April 1, 2020 (Census Day) and deliver state population
counts to the President by December 31, 2020 in order to determine the
Leveraging In-Person
number of congressional seats apportioned to each state. The Bureau is
Data Collection in Order to also required by law to deliver population counts to the states within 1
Meet Statutory Deadlines year of Census Day for redistricting purposes—March 31, 2021, in the
case of the 2020 Census. Even in normal circumstances, conducting the
count to meet these deadlines is an enormous and complex undertaking,
for which the Bureau spends years planning. This decennial, the COVID-
19 pandemic made the effort all the more challenging for the Bureau.
Managing in-person contact was one of the challenges of conducting the
census during a pandemic. Some data collection methods allowed
census staff to avoid such contact. For example, people could self-
respond to the census via the mail, phone, or internet. The Bureau also
collected data on household occupancy and other characteristics through
administrative records it has determined to be of sufficiently high quality
from agencies like the Internal Revenue Service and U.S. Postal Service.
However, for a significant part of census data collection, census
enumerators have to engage in in-person contact, in order to collect
responses. Much of this work was designed to be conducted in spring
2020 (see table 1).
Page 2 GAO-21-142 2020 Census
Table 1: Several 2020 Census Operations Relied on In-Person Contacts by Design
Planned number of Original
2020 Census field operationsa enumerators schedule
Non-Response Follow-Up 435,000 May 13, 2020–
Door-to-door interviews to collect data from households that have not yet responded July 31, 2020
to the census.
In-Person Group Quarters Enumeration 69,000 April 2, 2020–
Enumerate people who live in college and university dorms, correctional facilities, June 5, 2020
and skilled nursing facilities, among other facilities.
Service-Based Enumeration March 30, 2020–
Enumerate people in facilities such as transitional shelters, food kitchens, and non- April 1, 2020
sheltered outdoor locations that are primarily populated by people experiencing
homelessness.
Update Leave 24,000 March 15, 2020–
Enumerators update address information and deliver census questionnaires for April 17, 2020
respondents to return in rural, remote, or sparsely populated areas of the United
States and Puerto Rico, among other places. If a resident answers the door, the
enumerator conducts a brief interview to verify the address information.
Source: U.S. Census Bureau (Bureau) planning documentation. | GAO-21-142.
a
The Bureau also conducted smaller field operations to update address information and enumerate
households in remote locations, both of which involved in-person contact.
The Bureau also set up outreach efforts to populations identified as hard-
to-count that would include in-person activities, such as at community
information sessions and fairs. For example, between February 2020 and
May 2020, the Bureau scheduled outreach events in several major U.S.
cities including events that focused on specific hard-to-count populations,
such as young children.
Additionally, through its Mobile Questionnaire Assistance (MQA)
operation, the Bureau worked in early 2020 to identify key locations such
as houses of worship, community festivals, public transit hubs, libraries,
community centers, and other locations where people naturally
congregate to promote and assist with responding to the census. Through
this and other community engagement efforts, the Bureau hoped to
improve its outreach to and enumeration of hard-to-count population
subgroups, including young children, rural residents, and persons
experiencing homelessness. The Bureau first prioritized locations based
on historical data on self-response and later incorporated real-time self-
response data.
Page 3 GAO-21-142 2020 Census
COVID-19 Outbreak The outbreak of COVID-19, a strain of coronavirus to which the public
Overlapped with the does not have immunity, was first reported on December 31, 2019, in
Wuhan, China. 3 In the weeks that followed, the virus quickly spread
Bureau’s Scheduled Peak
around the globe, as did recognition of the virus’s severity. On January
Census Field Activity 31, 2020, the Secretary of Health and Human Services declared a public
health emergency for the United States, 4 retroactive to January 27, and
on March 11, 2020, the World Health Organization characterized COVID-
19 as a global pandemic. 5 In order to limit social contact and slow the
spread of the pandemic, states began implementing policies such as
stay-at-home orders that restrict travel and closures of nonessential
businesses.
As the early events of the COVID-19 outbreak unfolded in the United
States, the Bureau was just beginning to collect data for the decennial
census. On March 12, 2020, households began receiving the first of up to
five planned mailed invitations to respond to the census, and self-
response options were made available. At that time, the Bureau was also
enumerating residents in remote Alaska and was preparing to deploy field
staff for the Update Leave operation on March 15, 2020. Non-Response
Follow-Up (NRFU)—one of the most costly and labor-intensive decennial
field operations—was scheduled to begin nationwide on May 13, 2020. 6
3COVID-19 is caused by a new coronavirus named Severe Acute Respiratory Syndrome
coronavirus 2 (SARS-CoV-2). There are several different types of coronaviruses, some of
which are responsible for the common cold, and some of which cause severe respiratory
illness and have high mortality rates.
4A public health emergency triggers the availability of certain authorities under federal law
that enable federal agencies to take actions, such as temporarily reassigning certain state
and local personnel and waiving certain administrative requirements.
5Subsequently, on March 13, 2020, the President declared COVID-19 a national
emergency under the National Emergencies Act and a nationwide emergency under
section 501(b) of the Robert T. Stafford Disaster Relief and Emergency Assistance Act
(Stafford Act). The President has also approved major disaster declarations under the
Stafford Act for all 50 states, the District of Columbia, and five territories.
6The Bureau originally planned to enumerate college students living off-campus in April
2020 as part of what it called the early-NRFU operation, but this phase of NRFU was
eliminated amid the COVID-19 operational suspensions. The Bureau worked with colleges
remotely to collect data on these residents.
Page 4 GAO-21-142 2020 Census
Bureau Needs to
Examine Effects on
Data Quality of Its
Design Changes in
Response to COVID-
19
Bureau Revised Its As figure 1 shows, the changes to the Bureau’s operations took place
COVID-19 Response amid escalating COVID-19 case counts and state-level policy responses.
Multiple Times and
Ultimately Truncated Its
Field and Response
Processing Operations
Page 5 GAO-21-142 2020 Census
Figure 1: The Census Bureau Adjusted Its Operational Timeline during the COVID-19 Pandemic
The Bureau’s response to COVID-19 and the events that prompted them
took place over multiple phases.
Page 6 GAO-21-142 2020 Census
Initial suspensions. As part of its initial COVID-19 response and to
promote the safety of its workforce and the public, the Bureau delayed its
field operations multiple times. On March 18, 2020, the Bureau
announced it would suspend operations for 2 weeks. On March 28, 2020,
the Bureau announced another 2-week suspension. By the end of these
suspensions, the U.S. had reached over 600,000 total COVID-19 cases.
In recognition of the pandemic’s community spread, the Bureau on April
13, 2020, announced that it would extend the suspension of NRFU to a
total of 3 months. As a result, NRFU did not begin in earnest until August
9, 2020. The Bureau also adapted multiple operations as part of this
phase of response. Examples include efforts to encourage additional
facilities being contacted for Group Quarters enumeration to change to
more virtual methods of enumeration, as well as shifting community
partnership events and Mobile Questionnaire Assistance resources—
designed to encourage self-response—from planned social gatherings to
virtual events and less crowded physical locations like grocery stores and
pharmacies.
Plan for delayed delivery of population counts. Also on April 13, 2020,
the Bureau and Department of Commerce announced plans to deliver
population counts for apportionment purposes by April 30, 2021, and for
redistricting purposes by July 31, 2021—4 months later than the
respective statutory deadlines. According to the announcement, the
Department of Commerce had requested that Congress grant a 120-day
extension to the Bureau’s reporting deadlines. The Bureau also
announced an extension to the self-response period and plans to conduct
the NRFU operation through October 31, 2020.
This plan gave the Bureau 6 months from the end of NRFU to finish
processing data to prepare the apportionment counts. This was 1 month
longer than the Bureau had initially planned, but Bureau officials told us
that the extra month was designed as a contingency for further potential
disruptions to field operations, such as a second wave of COVID-19.
Gradual resumption of operations. The Bureau first resumed a revised
Update Leave operation on May 4, 2020, and announced a phased
approach to starting NRFU. 7 Between July 16 and August 9, 2020, the
7During the resumption of operations, the Bureau instructed enumerators to no longer
interview respondents about their address information when delivering questionnaires
during Update Leave in order to minimize possible in-person contacts.
Page 7 GAO-21-142 2020 Census
Bureau deployed three waves of a “soft launch” of NRFU by area census
offices (ACO) in areas that the Bureau deemed sufficiently safe. 8 The
Bureau selected initial ACOs using, among other criteria:
• the 2-week per capita COVID-19 case trends for all counties with
greater than 20 percent land area or housing units within a given
ACO’s boundaries,
• the ability of the Bureau to deliver personal protective equipment to a
given ACO’s field workforce, and
• the availability of sufficient field staff (enumerators and census field
supervisors) to do the work.
According to the Bureau, this soft launch approach was intended to
ensure that operational systems and procedures were ready for
nationwide use. The Bureau reported that the last ACOs started NRFU by
August 9.
Timeframes revised again to meet original statutory deadlines. On
August 3, 2020, the Bureau announced that it would accelerate its
operational time frames, as directed by the Secretary of Commerce, to
deliver population counts for apportionment and redistricting purposes by
the statutory deadlines, 4 months earlier than the plan the Bureau was
implementing. To meet the revised timeline, the Bureau announced that
NRFU and the self-response period would now end on September 30,
2020—1 month earlier than previously announced. Bureau officials stated
that streamlined post-data collection processing would account for the
remainder of the necessary timeline compression.
The Bureau also announced a series of operational changes to meet the
new timeframes. During the final 2 planned months of NRFU,
enumerators received awards for working more hours in the field and
were also given the option in most places of making phone calls to collect
data instead of making personal visits. Additionally, the Bureau
implemented a revised NRFU contact strategy to make fewer-than-
originally anticipated visits to households self-reported as vacant on
Census Day. The Bureau also temporarily reduced the scope of its NRFU
re-interview operation and compressed the time that subject matter
experts have to review apportionment population counts.
8The Bureau implemented field operations through a nationwide network of 248 ACOs,
which oversee local recruiting and operations and are organizational sub-units of offices in
the Bureau’s six regions.
Page 8 GAO-21-142 2020 Census
In September 2020, however, the U.S. District Court for the Northern
District of California issued a preliminary injunction, staying the August
2020 design changes, enjoining the Bureau’s adherence to the statutory
deadlines, and requiring the Bureau to continue data collection through
the end of October. 9 On October 13, 2020, the Supreme Court granted a
stay allowing the Bureau to cease data collection, while the appeal of the
preliminary injunction continues to be reviewed by the United States
Court of Appeals for the 9th Circuit. 10
Data Quality Concerns We have previously noted that late design changes increase the risks to a
Remain for the Bureau to quality census. The Bureau made operational changes in response to
COVID-19 and subsequent developments that include shortening the time
Address during
frames for implementing operations, making fewer visits to certain non-
Assessment of Census responding households, and reducing the scope and time of quality
checks on field data collection and data processing, respectively.
For its part, the Bureau has multiple mechanisms to assess the quality of
the 2020 Census. For example, through its Data Quality Executive
Governance Group, the Bureau oversees several teams of officials
dedicated to studying the effects on quality of the initial COVID-19
response, each with a set of objectives or deliverables that they are
planning to facilitate retrospective determinations of data quality. In
response to recommendations and input from entities like the American
Statistical Association, the Bureau is also working with a group of
independent scientific consultants known as the JASON group to assess
Bureau plans, processes, procedures, and metrics around 2020 Census
data quality and help develop a near-term strategy for sharing indicators
of the quality of census data that is produced for apportionment and
redistricting purposes, according to Bureau draft documentation. 11
9National Urban League v. Ross, Case No. 20-cv-05799 (N.D. Cal. Sept. 24, 2020) (order
granting motion for stay and preliminary injunction). The Department of Justice, on behalf
of the Department of Commerce and the Bureau, appealed the order. National Urban
League v. Ross, Case No. 20-16868(9th Cir., Sept. 25, 2020). When the 9th Circuit denied
a stay of part of the preliminary injunction, the ruling was appealed to the U.S. Supreme
Court.
10Ross v. National Urban League, Case No. 20A62, 592 U.S.___ (2020).
11The American Statistical Association and Census Scientific Advisory Committee issued
numerous recommendations in the Fall of 2020, including for the Bureau to document
what it knows in near real-time about the quality of the population counts that it provides to
the President and to Congress.
Page 9 GAO-21-142 2020 Census
In addition to the efforts described above, the Bureau also has numerous
planned assessments and evaluations of operations which, in conjunction
with its post-enumeration survey (PES)—a survey conducted
independently of each census to determine how many people were
missed or counted more than once—help determine the overall quality of
the census and document lessons for future censuses. The Bureau also
plans to release in mid-December data generated by its Demographic
Analysis effort, which is a specific set of techniques for developing
national population estimates by age, sex, and race from administrative
records including birth, death, and migration patterns. This will serve as
an initial independent check on census counts.
The Bureau is still updating its plans for these efforts to examine the
range of operational modifications made in response to COVID-19,
including the August 2020 and later changes. For example, according to
officials, the Bureau is revising its study plan template for its operational
assessments, which were first designed before the onset of COVID-19, to
include research questions that examine the effects of the COVID-19
response. The Bureau also indicated in late-August 2020 that the data
quality teams that are working in part to inform those assessment efforts
would be working through December 2020 to update the topics they are
monitoring to reflect any relevant August 2020 design changes.
Standards for Internal Control in the Federal Government indicate that
agencies should analyze and respond to identified risks, including by
considering relevant changes that have taken place within the agency’s
environment. 12
In preparing this report, we identified a number of concerns for the
Bureau’s assessments to address regarding how late changes to the
census design could affect data quality. Addressing these concerns as
part of updating the Bureau’s overall assessment of the quality of the
2020 Census will help to ensure public confidence and help the Bureau
begin to plan for 2030.
Initial Suspensions Due to • Recall bias could affect census responses. The reference date for
COVID-19 the census is April 1. Yet, the Bureau ultimately continued NRFU and
self-response through October 15, 2020. Recall bias—in this case, the
ability to remember where one was living and how many people
12GAO, Standards for Internal Control in the Federal Government, GAO-14-704G
(Washington, D.C.: Sep. 10, 2014).
Page 10 GAO-21-142 2020 Census
resided at a given address on April 1—has the potential to affect
responses the farther removed from April 1 that they are.
• Delays in enumerating persons experiencing homelessness
could affect the population count. Service-Based Enumeration
(SBE) is a primary operation for counting persons experiencing
homelessness and receiving services at facilities. Unlike for other
components of the census, the Bureau conducts SBE at a point in
time without a reference date due to the transitory nature of the
population. This operation was originally scheduled for the end of
March 2020—in proximity to Census Day. Due to COVID-19,
however, the Bureau delayed this operation until late September
2020.
The Bureau told us that this decision was made in consultation with
stakeholder groups and balanced the need for public safety amid the
pandemic with closely mirroring the climate conditions of when SBE
was originally scheduled to take place.
However, the economic recession that took place due to the
pandemic could have implications for this portion of the census.
According to the U.S. Bureau of Labor Statistics, at the end of March
2020 the unemployment rate was 4.4 percent, while at the end of
September 2020 the rate was 7.9 percent. Estimates from the Bureau
of Economic Analysis also show that Gross Domestic Product
decreased at an annualized rate of 31.4 percent during the second
quarter (April through June) of 2020. To the extent that such a drastic
economic downturn could affect the rate of homelessness, the
delayed operation has the potential to include responses of individuals
who had not been experiencing homelessness in March 2020 and
may have already responded to the census. 13
Because people experiencing homelessness are likely to have
transient living arrangements, the Bureau’s pre-COVID 19 plan of
conducting SBE in proximity to April 1 would have furthered the
overall goal of counting people in the right location as of April 1.
Individuals counted in September 2020 may be at different locations
than in March 2020, thus misrepresenting what the geographic
13The Bureau has procedures to de-duplicate census responses in cases when multiple
responses exist for the same person. However, we observed during testing that it is
possible for responses collected during SBE to not have sufficient information about the
respondent to allow for such de-duplication.
Page 11 GAO-21-142 2020 Census
distribution of the population was near census day. The Bureau told
us that the counts coming from SBE are a small percentage of the
total national count and that they expect the delay will have a minimal
effect on the overall census, but they did not explain how, if at all, they
plan to document any resulting possible limitations on the count of this
segment of the population within other totals in the data products. 14
• Operational delays could affect coverage measurement. The
Bureau is still revising its field data collection methodology for its PES
to determine how many people were missed or counted more than
once. As with other decennial field operations, the PES involves the
potential for person-to-person contact during enumeration, which
raises the risk that, without the Bureau changing data collection
procedures, people may not want to participate, affecting the quality of
the survey’s measurements.
Moreover, because deployment of the enumeration phase of the PES
was also delayed as part of the Bureau’s response to COVID-19, PES
is also susceptible to increased recall bias as discussed above. The
Bureau told us that its Demographic Analysis efforts were not affected
by COVID-19.
Acceleration of Field • The compression of NRFU timeframes creates uncertainty for the
Operations quality of NRFU across geographic areas and population
subgroups. There is a risk the data collected by the Bureau will be
less complete due to changes the Bureau made to accelerate NRFU,
at the Department of Commerce’s direction. By early September, the
time of the first of a series of court rulings that ultimately allowed the
Bureau to end NRFU on October 15 (two weeks early), the Bureau
reported completing over 68 percent of its NRFU workload, meaning
that it had conducted most of the operation under procedures
designed to meet an accelerated September 30 deadline (over four
weeks early). Some of these procedures, such as offering pay
bonuses to active enumerators for working more hours and exceeding
a specified level of completed NRFU cases per hour, had not been
previously tested, so the effects on the quality of data collected by
NRFU, if any, are unknown.
While the Bureau reported high completion rates nationwide, national
metrics can obscure important local variations. For example, on
14In addition to the apportionment and redistricting data products, the Bureau also
produces a series of demographic and housing data products directly using 2020 Census
data.
Page 12 GAO-21-142 2020 Census
September 27 (covering data through September 26), when the
Bureau reported that it was 95 percent complete nationally, it also
reported that 14 of the 248 ACOs were at less than 85 percent
complete. Further, two ACOs, Shreveport, Louisiana and Window
Rock, Arizona, were below 75 percent complete at that time. 15
While data quality effects nationally and within individual population
subgroups, if any, due to changes to accelerate NRFU are not yet
known, gaps in census data have historically disproportionately
affected certain population subgroups, such as non-Hispanic Blacks,
Hispanics, and American Indians on reservations. The Bureau has
long-standing methods to impute, or statistically derive, the numbers
and characteristics of residents of non-responding households using
information on nearby households. While the Bureau reported getting
census responses for over 99 percent of households by the end of
NRFU, it is not yet known the extent to which imputation will be
needed to determine missing demographic characteristics.
The Bureau told us that these methods of characteristic imputation
were not designed to be applied to large segments of the population.
Given these factors, if efforts to accelerate NRFU had any effects on
data quality, there is the question of whether there was
disproportionate impact on the quality of data for historically
undercounted groups. The Bureau acknowledges that its PES will
need to assess whether areas that remained relatively behind in
NRFU production also contributed to any differential undercounts
across population subgroups.
• Increased reliance on sole-source administrative records data in
some cases could affect the count. As one of its major cost-saving
innovations for the 2020 Census, the Bureau used administrative
records from sources such as Medicare enrollment data and individual
tax return data to determine housing unit occupancy as appropriate,
and, if necessary, enumerate non-responsive households. Preliminary
Bureau data show that administrative records helped enumerate or
15Our forthcoming December 2020 cost and progress update on the 2020 Census will
provide final data on NRFU completion across the country and ACOs. For examples of
prior installments in this series, see GAO, 2020 Census: Recent Decision to Compress
Census Timeframes Poses Additional Risks to Census Count, GAO-20-671R
(Washington, D.C.: Aug. 27, 2020); 2020 Census: COVID-19 Presents Delays and Risks
to Census Count, GAO-20-551R (Washington, D.C.: June 9, 2020); and 2020 Census:
Initial Enumeration Underway but Readiness for Upcoming Operations is Mixed,
GAO-20-368R (Washington, D.C.: Feb. 12, 2020).
Page 13 GAO-21-142 2020 Census
resolve as unoccupied nearly 14 percent of the households eligible for
NRFU, thereby reducing costly field visits and the need for imputation.
The Bureau modified its plans for using some administrative records
to complete the count rather than relying on potentially higher
numbers of interviews with neighbors or other proxies for
nonresponding housing units when compressing the NRFU timeline.
Specifically, the Bureau’s original plan had been to use a given source
of administrative records to determine the population of a given
household only when corroborated by another source of
administrative records. However, facing the possibility of higher rates
of incomplete cases or reliance on data from proxies, which the
Bureau considers generally as less reliable than other forms of data
collection, the Bureau decided late during NRFU to enumerate some
cases using sole sources of administrative records when available.
While the Bureau noted that this change would reduce the NRFU
workload for households that lack a population count, we have
previously reported on limitations with certain sources of
administrative records. 16 For example, Internal Revenue Service
individual tax return data can exclude people who have insufficient
income to file tax returns and could exclude residents who are not
listed as dependents for tax purposes. A tax filer may also include
dependents for tax purposes who do not live at the household’s
address full-time. Relying on tax return data for population counts
without corroboration could introduce related errors to household-level
census data.
The Bureau told us that how they determine the fitness for use of
administrative records for any given household sources did not
change during NRFU. The Bureau is reviewing the need to use
administrative records during post-processing in hurricane-affected
areas of Louisiana. Additionally, according to Bureau officials, they
are mitigating the possibility that some of the sole source data may be
missing young children, for instance, by linking any partial information
16GAO, 2020 Census: Bureau Is Taking Steps to Address Limitations of Administrative
Records, GAO-17-664 (Washington, D.C.: July 26, 2017).
Page 14 GAO-21-142 2020 Census
the Bureau has about a household to a database on children it has
created from administrative records. 17
The Bureau also said that while they will be examining quality effects
of the late design changes to application of administrative records, the
risks should also be mitigated by the fact that the Bureau will only be
applying this change to households that lack a population count
during the closeout stage of NRFU, after most households have
already been counted.
Streamlining of Response • The compressed timeframes for quality reviews and systems
Processing testing could affect the quality of the apportionment data file. The
Bureau’s August 2020 design changes reduced the time between the
end of NRFU and the provision of population counts for apportionment
purposes from 153 days (as planned following the COVID-19
suspensions) to 92 days, which was further reduced to 77 days
following court proceedings.
To meet this reduced time period, the Bureau streamlined reviews by
subject matter experts of the accuracy of the state-by-state counts.
According to officials, in August 2020 the Bureau did not remove any
levels of review when reducing the time to complete operations, but
as of October 2020 the Department of Commerce had not provided us
with copies of the Bureau’s streamlined processing timeline.
According to Bureau documentation, the Bureau had to re-process
census population files state-by-state for many states during prior
decennial censuses when Bureau demographers uncovered critical
errors during these reviews. Specifically, during post-data collection
processing for the 2010 Census, all states had their processing files
re-run at least once prior to apportionment tabulation, and 33 states
had their files re-run at least twice. Bureau documentation indicates
that there is little or no time left in the current streamlined processing
schedule to reproduce files, let alone research and identify any root
causes of errors that may be found.
The Bureau also completed testing for the 12 information technology
(IT) systems needed to support response processing in October 2020
17GAO-17-664. As we reported in 2017, the Bureau has created a database that merges
several administrative record sources to identify relationships between household adults
and children.
Page 15 GAO-21-142 2020 Census
instead of January 2021, as previously planned. 18 This change
occurred after the Bureau’s August 3, 2020, announcement that it
would accelerate its operational time frames to deliver population
counts for apportionment purposes.
As a result, there was less time available to complete system
integration testing and operational testing needed before response
processing began. The shortened time for response processing
discussed earlier added additional risk because there will have been
less time available to address system defects or other issues. We
have ongoing work intended to monitor the risks to the Bureau’s
remaining implementation of IT to support the 2020 Census and its
efforts to mitigate these risks.
• Compressed timeframes could limit the data available to review
for apportionment purposes. Bureau officials told us that, to enable
the compressed time frame for post-data collection processing, they
had to create a locked-down benchmark of the Master Address File
(MAF) of all living quarters by September 27, 2020—prior to the end
of NRFU and the self-response period. New addresses detected
would not become part of the MAF for apportionment or redistricting
purposes.
Bureau officials told us that the nature of field work is such that many
addresses that get added during the end of NRFU are usually variants
of existing addresses and thus do not contain new potential
respondents. Bureau officials indicated that prior assessments did not
determine how many addresses were added during the final weeks of
NRFU or what their characteristics were in prior censuses.
• Uncertainty over disclosure avoidance plans could introduce
risks. The Bureau plans to implement a disclosure avoidance
technique, known as differential privacy, to its publicly released
statistical products to protect the confidentiality of its respondents and
18The Bureau is utilizing 52 systems to conduct the 2020 Census, 12 of which are needed
to support response processing. We have previously reported on the Bureau’s plans to
deploy the 52 systems multiple times across 16 operational deliveries, which includes
operations such as NRFU and response processing. For more information, see, for
example, GAO-20-671R; GAO-20-551R; and GAO-20-368R.
Page 16 GAO-21-142 2020 Census
their data. 19 These products include detailed race tables,
demographic profiles, and demographic and housing characteristics.
The Bureau plans to issue fit-for-use redistricting products—which are
expected to be the first data products that utilize differential privacy
protections—by the March 31, 2021, statutory deadline. 20
As of October 2020, the Bureau had not made changes to its plans
and schedules for disclosure avoidance for the redistricting products
to meet that deadline, but Bureau officials stated that this could
change given the uncertainty in the near-term data collection and
processing schedules.
In addition, officials in the Directorate for Research and Methodology
reported that, because of compressed time frames, they did not plan
to produce other data products (such as the Demographics and
House Characteristics products) until sometime after March 2021. The
officials reported that they do not yet have a schedule for developing
these other data products, also because of uncertainty about the
near-term schedule.
In September 2020, the Bureau’s Census Scientific Advisory
Committee described a level of uncertainty related to the Bureau’s
implementation of disclosure avoidance, and noted that the Bureau’s
schedule had been ambitious. 21 Thus, the committee recommended
that the Bureau delay the release of data products (including fit-for-
use redistricting products) until after the apportionment count to allow
time to perform more analysis on the Bureau’s disclosure avoidance
plans.
• Time constraints call into question the ability to provide
redistricting data by statutory deadlines. In September 2020, the
19Differential privacy is a disclosure avoidance technique aimed at limiting statistical
disclosure and controlling privacy risk. According to the Bureau, using differential privacy
means that publicly available data will include some statistical noise, or data inaccuracies,
to protect the privacy of individuals. Differential privacy provides algorithms that allow
policy makers to decide the trade-off between data accuracy and privacy.
20Redistricting products are used, among other things, to delineate voting districts by
state. 13 U.S.C. § 141(c).
21The Census Scientific Advisory Committee was created to advise the Bureau’s Director
on the full range of Census Bureau’s programs and activities, and provide formal review
and feedback on internal and external working papers, reports, and other documents
related to the design and implementation of census programs and surveys.
Page 17 GAO-21-142 2020 Census
Bureau announced that it would produce population counts for
redistricting purposes by the original statutory deadline of March 31,
2021—four months earlier than what the Bureau had been planning
since the initial COVID-19 suspensions—though it has provided no
timeline for doing so.
Producing detailed tabulations for redistricting involves many steps.
As we reported in December 2019, the Bureau must resolve
incomplete and conflicting information within individual household
responses. 22 According to Bureau data, 13 percent of the people
counted in 2010 had missing or conflicting information on person
characteristics (such as age, sex race, and whether the housing unit
is owned or rented) that the Bureau had to resolve during this
process.
The Bureau also has to review the address file for enumerated
persons at a much more granular level of geography than simply
state-by-state counts (e.g., census blocks). If the Bureau has to
adhere to an accelerated timeframe for creating this data, it is unclear
whether officials will have time to assure quality.
Lessons from Developing Some design changes from August 2020 could be helpful in future
and Implementing Late censuses. The Bureau implemented numerous design changes late
during NRFU in order to accelerate completion of the operation. If
Design Changes Could Be
forthcoming Bureau assessments and evaluations are able to reveal
Helpful for 2030 Planning which, if any, of these changes improved efficiency without introducing
unacceptable risks to the quality of census data, then those innovations
might also be worthy of considering for future decennials. If, on the other
hand, any changes are shown to have introduced unacceptable risk or
harmed data quality, then the Bureau will know better what to avoid in the
future.
Learning from how the Bureau has attempted to remain agile in the
face of external factors prompting late changes to the operational
plan. The Bureau relied on a number of governance structures it already
had in place to help react to COVID-19 developments and other demands
for late design changes. For example, it expanded the scope of
preexisting quality teams to begin to examine the potential effects of the
initial COVID-19 response on census operations and data quality.
22GAO, 2020 Census: Changes Planned to Improve Data Quality, GAO-20-282
(Washington, D.C.: Dec. 20, 2019).
Page 18 GAO-21-142 2020 Census
In addition, the Bureau used its preexisting fusion center concept to
translate streams of data on COVID-19 incidence into decisions about
field operations, such as when to reopen specific census offices after
suspending activity during the pandemic. 23 The evolving nature of the
pandemic has forced the Bureau to be agile in its decision-making, and
external interest in its decisions has subjected it to high levels of scrutiny.
The culmination of this experience has the potential to highlight for the
Bureau which of its governance structures or management processes
worked well and which, if any, may have been less effective or redundant.
Pre-existing Contingency Before the COVID-19 outbreak, the Bureau had an estimated $2.03 billion
Funds Appear Sufficient to in total unobligated contingency funds for 2020 operations. According to
documentation from August 2020 and subsequent discussion with
Cover Implementation of
officials, the Bureau anticipates that these contingency funds will be more
the Bureau’s COVID-19 than sufficient to address both the initial COVID-19 response and the
Response August 2020 design changes. Prior to implementation of the August 2020
design changes, the Bureau estimated that obligations to implement the
initial COVID-19 suspensions of operations would account for all but $994
million of that contingency funding, and with the August 2020 design
changes added, the Bureau would still have at least $187 million in
contingency funding.
This means that, in the event that the Bureau is not authorized to extend
the statutory deadlines for data products, the total cost of the Bureau’s
COVID-19 response and August 2020 design changes is estimated at
over $1.8 billion of its contingency funds. Examples of major categories of
new spending from what the Bureau had planned pre-COVID-19 include:
• Service dates for area census office (ACO) staff. The Bureau
retained its core staff at the 248 ACOs following the 3-month
suspension of operations during COVID-19. The Bureau estimated
that extending salaries for these employees during this time would
cost $195 million above the Bureau’s original plan. The Bureau also
estimated that paying temporary leave to enumerators who became
exposed to COVID-19 during operations would add another $90
million in costs.
• Outreach and partnership activities. Before COVID-19, the Bureau
had planned to finish its 2020 paid advertising campaign in June 2020
and cease other partnership and outreach activities at the planned
23The Bureau designed a fusion center to serve as a centralized management process for
responding to disruptions to the 2020 Census.
Page 19 GAO-21-142 2020 Census
end of NRFU in July 2020. Following the COVID-19 suspensions, the
Bureau estimated that extending these activities, as well as
distributing an extra mailing to encourage self-response, would add
$160 million in costs.
• Changes to IT contracts and additional devices for field staff. The
Bureau is relying on significant IT contractor support to conduct the
2020 Census. Specifically, the Bureau is utilizing five major IT
contracts to provide critical capabilities needed to conduct the census,
including call center support; handheld device procurement and
provisioning; and system and infrastructure development, testing,
deployment, and cybersecurity services. For example, the Bureau’s
Decennial Device-as-a-Service contract provided handheld device
procurement, provisioning, and distribution for 2020 Census field data
collection operations, such as NRFU.
Due to the COVID-19 pandemic, the Bureau stated that it made
substantial changes to its five major IT contracts, such as extending
software licenses, purchasing additional handheld devices, and
increasing the number of IT support staff. In total, as of October 2020,
Bureau officials stated that these changes had a cost impact of
approximately $357 million. Table 2 provides a description of each of
the five major IT contracts and examples of changes made to the
contracts to address COVID-19 impacts.
Table 2: Examples of Changes to 2020 Census Information Technology (IT) Contracts Due to COVID-19
Example of changes the U.S. Census Bureau made to
Major IT contract Description address COVID-19 impacts
Technical Integrator Provides 2020 Census IT infrastructure • Extension resulted in the contractor extending the
and systems development, testing, employment of technical integrator staff.
deployment, and cybersecurity services. • Extended the term of equipment licenses and
services, such as Amazon Web Services.
Decennial-Device-as-a- Provides mobile device procurement, • Purchased an additional 125,000 mobile phones and
Service provisioning, and distribution for 2020 5,500 tablets.
Census field data collection operations. • Extended mobile device management capabilities,
cellular services, and IT support services.
Census Schedule A Human Provides 2020 Census IT recruiting, • Increased the number of help desk support staff to
Resources and Recruiting, selection, and learning management support recruiting and assessment services.
Payroll System - Recruiting services. • Purchased additional licenses for software that
and Assessment and supports recruiting and assessment services.
Learning Management
System
Field IT Deployment Provides IT infrastructure and • Increased the number of staff and extended IT
telecommunications for 2020 Census field support hours for Paper Data Capture Centers and
sites, including Paper Data Capture Area Census Offices.
Centers and Area Census Offices.
Page 20 GAO-21-142 2020 Census
Example of changes the U.S. Census Bureau made to
Major IT contract Description address COVID-19 impacts
Census Questionnaire Provides call center capability to assist • Increased call center staff support from March 2020
Assistance respondents with completing the 2020 through October 2020.
Census questionnaire through the internet
self-response and phone response
options.
Source: GAO analysis of U.S. Census Bureau information. | GAO-21-142
• NRFU production awards program. To increase the number of total
hours worked and enable the Bureau to accelerate NRFU by a month
as part of the August 2020 design changes, the Bureau developed a
program that would award enumerators and census field supervisors
for exceeding Bureau targets for weekly hours worked and
productivity. These awards would escalate for employees who met the
eligibility requirements for multiple weeks over a period of time.
Overall, the Bureau anticipated this program adding over $320 million
in costs.
As of the time of this report, some costs associated with the August 2020
design changes are unknown. For instance, the Bureau may have to pay
certain employees overtime in order to implement compressed response
processing operations. Officials told us that any effects of these costs on
contingency funds would only become apparent if they exceeded what
had previously been budgeted for overtime for those activities.
Officials also noted that onboarding and training costs for additional
enumerators recruited as a result of the need to accelerate NRFU were
likely to be lower than the $229 million that the Bureau allotted. Officials
noted that this is because the Bureau did not need to replace as many
enumerators from attrition as expected. Finally, when announcing the
Bureau’s plan to abide by the September 2020 court ruling on the August
2020 design changes, the Bureau offered a $500 award to enumerators
who would travel to help complete NRFU in states that were experiencing
production shortfalls, though at the time of this report it is unclear how
many enumerators accepted this offer. 24
Conducting a census amid a global pandemic posed unprecedented
Conclusions challenges to the Bureau. Major events, ranging from a global pandemic
to multiple directed revisions to operational timeframes, have affected the
way that the Bureau has implemented the 2020 Census. Given these
24Places that the Bureau targeted for this award included the Window Rock reservation for
the Navajo Nation, as well as the states of Alabama, Georgia, Louisiana, Mississippi,
Montana, and South Carolina.
Page 21 GAO-21-142 2020 Census
disruptions, it will be critical for the Bureau to ensure that it is learning as
much as it can from the effects of its design changes. We have identified
data quality concerns prompted by these late design changes, such as
whether they will affect historically undercounted population subgroups or
whether the Bureau will have sufficient time to correct errors found in the
data. There may also be efficiencies or other benefits from innovations
the Bureau made out of necessity that it can replicate in future censuses.
The Bureau is updating planned operational assessments and
evaluations for the 2020 Census, as well as its post enumeration survey.
As the Bureau formalizes its updated approach, it will be important that it
carefully and thoroughly evaluate issues stemming from the operational
changes made in response to COVID-19. Doing so will help ensure public
confidence in the 2020 Census and inform future census planning efforts.
The Secretary of Commerce and the Director of the U.S. Census Bureau
Recommendations for should update and implement assessments, evaluations, and coverage
Executive Action measurement efforts to address the effects of the Bureau’s response to
COVID-19 that we identified, including data quality concerns and potential
operational benefits from innovations. (Recommendation 1)
We provided a draft of this report to the Department of Commerce. In its
Agency Comments written comments, reproduced in appendix I, the Department of
and Our Evaluation Commerce agreed with our findings and recommendation. The Bureau
also provided technical comments, which we incorporated as appropriate.
We are sending copies of this report to the Secretary of Commerce, the
Undersecretary of Economic Affairs, the Director of the U.S. Census
Bureau, and the appropriate congressional committees. In addition, the
report is available at no charge on the GAO website at
http://www.gao.gov.
If you or your staff have any questions about this report please contact J.
Christopher Mihm at (202) 512-6806 or by email at mihmj@gao.gov or
Nick Marinos at (202) 512-9342 or by email at marinosn@gao.gov.
Page 22 GAO-21-142 2020 Census
Contact points for our Offices of Congressional Relations and Public
Affairs may be found on the last page of this report. GAO staff who made
key contributions to this report are listed in appendix II.
J. Christopher Mihm
Managing Director, Strategic Issues
Nick Marinos
Director, Information Technology and Cybersecurity
Page 23 GAO-21-142 2020 Census
List of Addressees
The Honorable Ron Johnson
Chairman
The Honorable Gary C. Peters
Ranking Member
Committee on Homeland Security and Governmental Affairs
United States Senate
The Honorable Jerry Moran
Chairman
The Honorable Jeanne Shaheen
Ranking Member
Subcommittee on Commerce, Justice, Science, and Related Agencies
Committee on Appropriations
United States Senate
The Honorable Carolyn B. Maloney
Chairwoman
The Honorable James Comer
Ranking Member
Committee on Oversight and Reform
House of Representatives
The Honorable Jose Serrano
Chairman
The Honorable Robert B. Aderholt
Ranking Member
Subcommittee on Commerce, Justice, Science, and Related Agencies
Committee on Appropriations
House of Representatives
The Honorable Joaquin Castro
House of Representatives
The Honorable Judy Chu
House of Representatives
Page 24 GAO-21-142 2020 Census
Appendix I: Comments from the Department
Appendix I: Comments from the Department of
Commerce
of Commerce
Page 25 GAO-21-142 2020 Census
Appendix II: GAO Contact and Staff
Appendix II: GAO Contact and Staff
Acknowledgments
Acknowledgments
J. Christopher Mihm, (202) 512-6806 or mihmj@gao.gov
GAO Contact Nick Marinos, (202) 512-9342 or marinosn@gao.gov
In addition to the contact named above, Ty Mitchell, Kate Sharkey, Jon
Staff Ticehurst (Assistant Directors), Devin Braun, Andrea Starosciak
Acknowledgments (Analysts-in-Charge), Mark Abraham, Ann L. Czapiewski, Alan Daigle,
Alexandra Edwards, Rob Gebhart, Lisa Hardman, Jason T. Lee, Kerstin
Meyer, Lisa Pearson, Farrah Stone, and Peter Verchinski, made key
contributions to this report.
(104496)
Page 26 GAO-21-142 2020 Census
The Government Accountability Office, the audit, evaluation, and investigative
GAO’s Mission arm of Congress, exists to support Congress in meeting its constitutional
responsibilities and to help improve the performance and accountability of the
federal government for the American people. GAO examines the use of public
funds; evaluates federal programs and policies; and provides analyses,
recommendations, and other assistance to help Congress make informed
oversight, policy, and funding decisions. GAO’s commitment to good government
is reflected in its core values of accountability, integrity, and reliability.
The fastest and easiest way to obtain copies of GAO documents at no cost is
Obtaining Copies of through our website. Each weekday afternoon, GAO posts on its website newly
GAO Reports and released reports, testimony, and correspondence. You can also subscribe to
GAO’s email updates to receive notification of newly posted products.
Testimony
Order by Phone The price of each GAO publication reflects GAO’s actual cost of production and
distribution and depends on the number of pages in the publication and whether
the publication is printed in color or black and white. Pricing and ordering
information is posted on GAO’s website, https://www.gao.gov/ordering.htm.
Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537.
Orders may be paid for using American Express, Discover Card, MasterCard,
Visa, check, or money order. Call for additional information.
Connect with GAO on Facebook, Flickr, Twitter, and YouTube.
Connect with GAO Subscribe to our RSS Feeds or Email Updates. Listen to our Podcasts.
Visit GAO on the web at https://www.gao.gov.
Contact FraudNet:
To Report Fraud,
Website: https://www.gao.gov/fraudnet/fraudnet.htm
Waste, and Abuse in
Automated answering system: (800) 424-5454 or (202) 512-7700
Federal Programs
Orice Williams Brown, Managing Director, WilliamsO@gao.gov, (202) 512-4400,
Congressional U.S. Government Accountability Office, 441 G Street NW, Room 7125,
Relations Washington, DC 20548
Chuck Young, Managing Director, youngc1@gao.gov, (202) 512-4800
Public Affairs U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, DC 20548
Stephen J. Sanford, Acting Managing Director, spel@gao.gov, (202) 512-4707
Strategic Planning and U.S. Government Accountability Office, 441 G Street NW, Room 7814,
External Liaison Washington, DC 20548
Please Print on Recycled Paper.
File and source
- File
- gao-21-142.pdf
- Size
- 2,001,379 bytes
- SHA-256
- 6de67e0a1180fbcba2ac17324fd8c0f5771c762f7ba977f44fed9611f7b5a2ee
- Our copy
- gao-21-142.pdf
- Original
- www.gao.gov