PREPARED STATEMENT OF
PRESTON JAY WAITE, DEPUTY DIRECTOR
US CENSUS BUREAU
The 2010 Census
Before the Subcommittee on Information Policy, Census, and National Archives
Committee on Oversight and Government Reform
24 April 2007
On behalf of the U.S. Census Bureau, Mr. Chairman, I would like to thank the members of the subcommittee for this opportunity to discuss the status and progress of the 2010 reengineered decennial census. We are approaching Census Day — it is now less than three years away — and the goals established earlier this decade are in sight. The goals of the 2010 reengineered census design are to: reduce operational risk; contain costs; improve the relevance and timeliness of the census long-form data; and improve accuracy. To accomplish these goals, we developed an innovative and integrated program which includes three main components:
Each of these components is integral to the 2010 decennial census program. Their goals are complementary and the success of each component supports the overall success and accuracy of the 2010 Census—which is our most important goal as we proceed to Census Day.
In fact, the 2010 Census is already underway. In February, we sent informational letters to every local government throughout the country, outlining our plans for the Local Update of Census Addresses (LUCA) program. This is an important operation, as well as an important partnership for the census, and supports our key goal to improve the accuracy of the census. Last month, in accordance with the Census Act, we submitted to Congress those subjects planned in the 2010 Census, which include gender, age, race, ethnicity, relationship, and whether you own or rent your home. The 2010 Census form is estimated to take less than 10 minutes to complete, making it one of the shortest and easiest to complete since the nation’s first census in 1790. Next year, we will submit the wording of these questions to Congress.
Progress on the Major Components of the 2010 Decennial Census Program
Many efforts over the past years will improve accuracy, including the American
Community Survey, which was fully implemented in 2005 for the household portion
of the sample and is replacing the decennial census long form. With a sample
size of approximately three million households per year, or 250,000 per month,
we will be able to provide timely data on a yearly basis to local communities.
Last year, we began collecting data in group quarters, such as jails, nursing
homes, and college dormitories, fulfilling our last major collection objective
in replacing the long form. In 2006, we delivered the first annual data from
the American Community Survey for the nation, every state, and every governmental
unit with a population of 65,000 or more. In addition, last year we released
a special product for the Gulf Coast area affected by Hurricane Katrina. This
product demonstrated the power of the American Community Survey in measuring
change, showing the changes affecting communities after that devastating natural
disaster.
This August we will begin releasing the next set of annual data, which also include data for group quarters, and we will continue to deliver these data every year. In 2008, we will provide data to communities with populations of 20,000 or more, and in 2010 we will deliver these data for census tracts—a full two years before traditional census long-form data would have been available. This not only fulfills the goal of the American Community Survey but also achieves one of the primary goals of the reengineered 2010 census. Providing annual socio-economic data to local communities means they will have more relevant information to evaluate trends and measure change more effectively. Prior to the American Community Survey, data users had to wait 10 years for new local area data. This 10-year data gap hampered the ability to measure key factors in assessing change. For instance, without annual information, education planners could not assess a shift in the demographic composition, potentially missing the growth of young families in a transitional neighborhood. With annual data from the American Community Survey, planners have more current information to make better decisions.
The second component of the 2010 reengineered census program is the mapping and geography component known as the MAF/TIGER Enhancements Program. By correcting and modernizing both the address list information and the TIGER mapping system, we will ensure greater geographic accuracy of the census and enhance our ability to conduct field operations in the 2010 Census, improving census coverage. Geographic accuracy is vital because the census must fulfill two principal requirements: 1) count every person living in America, once and only once, and 2) count every person at the correct address. Each address corresponds to specific geographies: a census block, census tract, place, county, and state. Ensuring the accuracy of the addresses helps guarantee the fair distribution of political representation and resources, as they are distributed according to geographies — states, cities, towns, census tracts, and blocks.
With the MAF/TIGER Enhancements Program, we are working with the private and public sectors to modernize and enhance the capabilities of the nation’s road map. Important objectives of the program include realigning the street centerlines in the TIGER mapping system in order to take advantage of GPS capabilities, modernizing the processing system, and expanding geographic partnerships. We are modernizing the MAF/TIGER processing system, replacing the homegrown system developed more than 25 years ago before the information and technology revolution, with a Commercial-Off-the-Shelf system.
We are working with the Harris Corporation to realign all the U.S. counties by 2008, in time for field operations to conduct address canvassing operations for the 2010 Census. Since 2003, we have completed the realignment for nearly 2000 of the nation’s 3,232 counties. By April 2008, we are scheduled to complete the remaining counties in time to conduct the address listing operations. These efforts will enhance our ability to conduct census field operations, including non-response follow-up, and will provide a more flexible integration with other operations. MAF/TIGER tells us where people are living, and not only furnishes us with a list of households to contact, but also provides a reasonable means of organizing our workload and the non-response follow-up operations. We want field representatives to work with accurate maps and to be able to effectively use the handheld devices. Automation is one of the most important elements of the planned improvements for the 2010 Census—we believe it will help contain the costs of field operations, reduce operational risk, and improve geographic accuracy.
We also developed a sustained, comprehensive testing program in preparation for the 2010 short-form only census—the third component of the overall 2010 reengineered census plan—incorporating both major mail-out and field operation tests. Throughout the testing program, we are striving to ensure that we meet the overall goals of the 2010 reengineered census, most importantly to reduce operational risk and improve the accuracy and coverage of the decennial census. We have conducted key tests each year. Beginning with a national mail-out test in 2003 we have studied alternative self-response options and contact strategies, as well as alternative presentations of the race and Hispanic origin questions. In 2005, we conducted a second mail-out test to assess issues such as coverage questions, residence rules, replacement questionnaires, and the design, layout, and wording of race and ethnicity questions and other short form content.
In addition to these tests, we conducted major field tests in 2004 and 2006. In 2004, we conducted a test in the Queens Borough of New York City and in southwest Georgia, focusing primarily on using new data collection technologies. In this test, we also determined we could successfully train enumerators in a short time to make use of the handheld computer devices, which are fundamental to the design of the 2010 Census. We conducted a second major field test in 2006 in Travis County, Texas and the Cheyenne River Reservation in South Dakota. We selected these sites because their characteristics allowed us to answer specific research questions and provided a final opportunity to test methods and technologies in the field before they are integrated in the Dress Rehearsal.
2010 Census Updates
We believe these testing efforts will help us improve the accuracy of the responses,
and thereby census coverage. We will take these experiences and the research
we conducted thus far into the field with the 2008 Dress Rehearsal. The Dress
Rehearsal is our last opportunity to ensure planned procedures and operations
will function as designed once they are integrated with one another. The sites
for the Dress Rehearsal are in San Joaquin County, California and in Fayetteville
and the counties in eastern North Carolina, near Fort Bragg. We have opened
the Local Census Offices (LCO) in both locations and started hiring. The address
canvassing operations begin next month. We only get one chance to take the census.
We need a true rehearsal because we will not conduct untried procedures during
the 2010 Census.
We especially are aware of potential risks as we get closer to Census Day and begin actual decennial operations such as LUCA. As I mentioned previously, we sent advance notification letters to every state, tribal, county, and local government throughout the country, in anticipation of sending the actual invitations later this summer. These letters describe the LUCA program and outline key activities. We already are conducting more than 600 LUCA workshops over the next few months. These workshops are intended to answer questions and encourage participation. LUCA is one of the most important partnerships of the census. It is through working with local governments that we often learn of new housing construction, demolitions, and conversions, as well as map updates, which are fundamental to the geographic accuracy of the census. In contrast to the LUCA program for Census 2000, we are contacting local governments and conducting LUCA updates prior to address canvassing. We believe this will reduce confusion and ensure greater accuracy.
Greater accuracy is the primary focus of all of our efforts. This is especially
true of the coverage improvement program. The coverage improvement program for
2010 incorporates lessons learned from previous censuses, as well as results
from our multi-year research, development and testing program. Much of the housing
unit coverage error in Census 2000 resulted from geocoding errors. We believe
we will see major improvements that will result from our efforts to update and
modernize—from realigning the street centerlines in our TIGER database
to using GPS-equipped handheld computers during the address canvassing and non-response
follow-up operations. We also believe coverage will be improved through our
effort to maintain the MAF over the decade (primarily through work with the
U.S. Postal Service), and by making a number of improvements to the LUCA program.
In addition, we have conducted extensive testing of ways to better explain our
residence rules so the people better understand who should be included or excluded
as members of their household for census purposes.
Related to this, we have also tested the use of coverage probes on the questionnaire
to help identify households with potential undercounts or overcounts that may
result from respondents not understanding the residence rules. Preliminary results
indicate these questions can help identify households with coverage errors that
we can resolve through follow-up. And as with other censuses, we will conduct
coverage measurement evaluations. The 2010 Census coverage measurement program
will not only assess the completeness and coverage of the census, but will also
provide valuable insights for future censuses.
The 2010 Census will be an important milestone for the Census Bureau. Even as we are incorporating the lessons learned from past censuses, we are embracing new approaches—the American Community Survey, the use of GPS-based technology and automation, the second mailing. And we are seeking the knowledge and experience from a greater range of partners.
In addition to LUCA, where we are engaging local governments to take advantage of their knowledge, we are also engaging the private sector to benefit from their knowledge and experience to make the decennial census program more efficient.
Automation is a key component for the 2010 Census. To successfully achieve this objective, we have enlisted private sector partners to help us build the census data collection infrastructure and to assist the Census Bureau with the integration of its components. Part of our efforts have centered on two major systems, the 2010 Decennial Response Integration System (DRIS) and the Field Data Collection Automation (FDCA) system. Both of these are large information technology contracts, totaling together over $1 billion. We believe our efforts with DRIS and FDCA will result in reduced operational risk and improved accuracy. The purpose of the DRIS contract, which was awarded in 2005 to Lockheed Martin Corporation, is to ensure accurate and protected collection and storage of Americans’ data whether by paper form, handheld computer, or telephone. We are confident of this approach based on our experience from Census 2000 when we partnered with the private sector to conduct data capture. We currently are implementing and testing this system in preparation for the 2008 Census Dress Rehearsal.
For the 2010 Census, we also intend to use automation to directly capture information
collected during personal interviews in non-response follow-up and other field
operations, eliminating the need for paper maps and address lists for the major
field data collection operations. This is a significant change from the paper-based
census field operations of Census 2000 and every other previous census. The
FDCA contract was awarded last spring to the Harris Corporation. It provides
automation resources to support field data collection operations, including
an integrated IT infrastructure, as well as support for handheld devices and
other aspects of the field activities.
This contract will also provide for the purchase of handheld devices and the
operating system. This custom-designed device will be used to collect information
from households, provide geographic support to the enumerators, and gather administrative
and payroll information for the Census Bureau. The estimated cost of each handheld
device is $400. This cost includes the operating system, the GPS receiver, cellular
antenna, dialup modem, fingerprint sensor, memory card, car and household charger
adapters, battery, and the stylus, as well as software licensing. An important
feature of the handheld computer is its security protection. The Census Bureau’s
ultimate priority is to protect the information we collect. The devices will
require two-factor authentication for access, including a thumbprint and a “password”
answer. It will “lockout” users after 15 minutes of non-use. Moreover,
while these devices will resemble other commercially available equipment, they
are being developed specifically for the 2010 Census, with software capabilities
limited to those required for the census.
Of the remaining contracts, we have released the Request for Proposals and are evaluating the vendor proposals. We will announce the award for the printing contract, which includes printing, labeling, and assembling 676 million public use forms (of which 400 million are questionnaires) next month. The Data Access and Dissemination System award will be announced later this year. Finally, we will announce the award of the communication contract in early fall. The first key deliverable of this contract will be an integrated communications plan in the first quarter of FY 2008.
The communications contract will be at the forefront of a multi-faceted, integrated effort to increase the mail response rate, reduce the differential undercount even further, and encourage cooperation during the non-response follow-up operations. As part of the overall communications strategy, we intend to incorporate the lessons and successes of Census 2000, when for the first time in history we improved the mail response rate and reduced the differential undercount. Both paid advertising and partnerships were instrumental to these successes. One of the most important lessons of Census 2000 was the need to engage state, tribal, and local governments sooner rather than later. To that end, we are already contacting key local governments to discuss the Complete Count Committee program, so if they wish to initiate such a program, they will have a head start on planning their effort. Partnerships, whether they are formed through local governments or through national, local, or community organizations, are key to the success of the census. These partnerships encourage participation and demonstrate the importance of the census to every community. Partners are often our best ambassadors in hard-to-count communities and can encourage participation.
The census ultimately depends on the trust and participation of every household. To reach these households entails the success of a complex series of operations, ranging from the accuracy of the MAF to well-targeted advertising, all occurring in sequence and on time. Your continued support is vital, especially now as the decennial census ramps up. It takes an enormous effort, as well as precision-planned coordination to open more than 450 local census offices throughout the nation, contact 130 million households, and count more than 310 million people. At the Census Bureau, our eyes are figuratively fixed on Census Day—yet we know the census has already begun. The Dress Rehearsal is underway, as is the 2010 LUCA program. Next year, we will conduct the Dress Rehearsal enumeration and we will open the 12 Regional Census Centers and the early LCOs as part of our efforts to ensure the field infrastructure is in place.
I believe these efforts are an important investment in the success of the census and in our nation’s future, as the census data will be used to allocate both power and federal funding over the course of ten years. Many communities depend on the census—the count, as well as the information about the population and housing—to make critical decisions that affect the lives of every person living in America.
Thank you, Mr. Chairman, and I would be happy to answer your questions.
Source: U.S. Census Bureau
Public Information Office
301-763-3030
