PREPARED STATEMENT OF
STEVE H. MURDOCK
DIRECTOR
U.S. CENSUS BUREAU
The Fiscal Year 2009 Budget
Before the Subcommittee on Commerce, Justice, Science, and Related Agencies
Committee on Appropriations, U.S. House of Representatives
3 April 2008
On behalf of the U.S. Census Bureau, Mr. Chairman, I would like to thank you and the members of the subcommittee for this opportunity to discuss our plans for the 2010 Decennial Census and the Census Bureau’s FY 2009 budget. We would like to thank the Congress, particularly the members of this subcommittee, for their support of the Census Bureau’s budget over the past several years. Your continued support will enable us to undertake a complete and accurate census.
Secretary Gutierrez has just outlined our plans for going forward with the Field Data Collection Automation (FDCA) program. Since the Secretary last spoke to you, all of the options regarding the FDCA program were thoroughly examined. We have considered feasibility, timing, and, of course, costs, while placing our greatest emphasis on ensuring the most accurate possible census in 2010. I fully support his decision. I believe this path will enable us to successfully complete the 2010 Census.
We now understand that the problem with the FDCA program was due to a lack of communication between the Census Bureau and the prime contractor for FDCA, and to difficulties the contractor had in developing the full scope of the project within our deadlines. From the beginning, we did not effectively convey to the contractor the complexity of census operations, and the detailed requirements that needed to be fulfilled in order to complete the operations that FDCA covers. Once these detailed requirements were completely delineated, we had serious concerns about rising costs and our ability to complete a successful 2010 Census if we continued developing the FDCA program as planned.
The FDCA program was designed to supply the information technology infrastructure, support services, hardware and software to support a network for several hundred local offices and more than 500,000 hand-held computers (HHCs) that would be used around the country. It is helpful to think of FDCA as being made up of three fundamental components:
Automated data collection using handheld devices both to verify addresses, called Address Canvassing, and to collect data during the nonresponse follow-up, known as NRFU, of those households that do not return the census survey;
The Operations Control System (OCS) that tracks and manages Decennial Census workflow; and
In late November 2007, as a result of concerns raised regarding the ability to meet deadlines and budgets, the Deputy Director of the Census Bureau initiated a comprehensive assessment to determine the status of the program and to better understand any issues or concerns as the program approaches key 2010 Census milestones. This assessment included a series of wide-ranging meetings with Census Bureau staff directly involved in the FDCA program. The Deputy Director also met with Harris Corporation, the company developing the FDCA system; and Mitre Corporation, an information technology firm under contract with the Census Bureau. Mitre’s role is to provide an internal, independent assessment of the information technology systems in the decennial programs and also IT systems in the Census Bureau.
This process identified issues that raised concerns about the complete development of all of the operations initially planned for the FDCA system in time for the 2010 Census. These included:
HHC Functionality -- Assessments of the Address Canvassing operation for the 2008 Dress Rehearsal revealed that there were difficulties in obtaining efficient transmission to and from the hand-held computer, resulting in enumerator downtime. In addition, the HHCs did not function well if the number of addresses in the enumerator assignment area was too large.
OCS Requirements -- The sheer volume of requirements, as well as the complexity of the operations that FDCA supports, contribute to problems that are particularly significant in the development of the OCS. Contract deliverables in the fall indicated that problems remained despite steps taken to clarify and strengthen the process we use to define contract requirements.
Accordingly, the Deputy Director established an Integrated Project Team (IPT) made up of key, high ranking 2010 Census managers. The IPT was tasked with producing the final set of FDCA program requirements by January 16, 2008.
This process was nearing completion when I arrived in early January. When Harris Corporation provided feedback at the end of January, the full scope of our problem came into focus.
As we grappled with this problem, I established a task force, chaired by former Deputy Director William Barron and made up of some of the Census Bureau’s and the Department’s best people, as well as representatives from Mitre, to help us develop a strategy to address these problems once and for all.
The Task force outlined four options for moving forward. All of these options call for using the handheld computers for Address Canvassing, and we are continuing to work to ensure this requirement is met. For the other major components of FDCA, each of the options considers a combination of responsibilities between the contractor and the Census Bureau in terms of capabilities, expertise, staffing, timing, and costs.
The work of the task force was then turned over to the Expert Panel established by the Secretary and made up of two former Census Bureau Directors, a former Associate Director of the Census Bureau, information technology experts, and a former Member of Congress.
After deliberating with this panel, the Secretary recommended the plan that he described in his testimony. I fully support the Secretary’s decision. Our contractor will continue developing the Address Canvassing operation utilizing the handheld computers and develop the Operations Control System, and the Census Bureau will implement a paper-based NRFU operation and provide the RCC infrastructure. We will work with our contractor to ensure that the other components of the FDCA program are successfully completed. This option maximizes our control of 2010 Census operations. The Census Bureau knows how to develop and implement a paper-based NRFU, and our decision to do so again gives us flexibility and minimizes the risks we identified in the FDCA program. At the same time, the plan allows us to leverage Global Positioning System technologies by using handheld computers in the Address Canvassing operation. This will improve the accuracy of our address list, which is fundamental to an accurate census.
Since becoming Director in January, addressing the problems associated with the FDCA program has been my highest priority. With the replan outlined today, I am confident we can have the 2010 Census back on track.
During this hearing I think it is also worthwhile to turn our attention to other operations underway to prepare for the 2010 Census, and to conduct the 2007 Economic Census and the demographic surveys that are central to the Census Bureau’s mission.
The decennial census is the highest priority within the Census Bureau and the request for over $2.1 billion accounts for over 80 percent of the original requested 2009 budget. As noted by the Secretary, we will need an additional $160 to $230 million in 2008, and we will be asking for approximately $600 million more in a budget amendment for 2009. The department has identified offsets to achieve the 2008 funding within departmental resources, provided it receives expanded transfer authority. The 2009 budget amendment will also be accompanied by offsets to be budget neutral.
The 2010 Census is underway. All 12 Regional Census Centers and the Puerto Rico Area Office are open for business. With funding provided by the FY 2008 appropriation, we have hired the first 48 partnership staff and will hire an additional 72 in May. We have provided over 11,000 participating governments with materials for the Local Update of Census Addresses (LUCA) program. LUCA plays a critical role in our efforts to achieve an accurate count and is one of our most important partnerships. By working with the local governments we learn of new housing construction, as well as demolitions and conversions, information fundamentally important to counting everyone.
The plans for the 2010 Census also include important structural improvements and enhancements to the Nation’s road map: the Master Address File (MAF) and the Topologically Integrated Geographic Encoding and Referencing mapping system, better known as TIGER. Each of these activities is interdependent; the success of each component contributes to the overall success of the 2010 Decennial Census. Specifically, the MAF/TIGER Enhancement Program is a multi-year effort to collect and correct the locations of streets and other geographic information. Important objectives of the program include realigning the TIGER map to take advantage of GPS capabilities, modernizing the processing system, and expanding geographic partnerships. For FY 2009, we are requesting $18 million for this program. We are working with our contractor to complete the effort to realign the street-centerlines for all 3,232 counties throughout the U.S. and its island areas by the end of FY 2008. In FY 2009, we will conduct closeout activities for the realignment contract and work to maintain the fruitful geographic partnerships we have established with many state and local governments, and also evaluate the success of the realignment effort.
This is vital because the census must fulfill two principal requirements: to count every person living in America, once and only once; and to count every person at the correct address. The MAF tells us where the housing units are located and not only furnishes a list of addresses to contact, but also provides a reasonable means of organizing our workload and the non-response follow-up and tabulation operations. Ensuring the accuracy of the addresses is an essential guarantee for the fair distribution of representation and resources, as they are distributed according to geographies — states, tribal governments, counties, cities, towns, census tracts, and blocks. In other words, the accuracy and the success of the census—both the short-form only decennial census and the American Community Survey (ACS)—ultimately depend upon the accuracy and completeness of the MAF and TIGER systems.
The success of the 2010 Decennial Census Program also depends upon the American Community Survey (ACS), the largest household survey in the United States, with an overall sample of approximately three million households per year or about 250,000 households per month. The ACS replaces the traditional decennial census long form data collection and will provide annual detailed socio-economic information for every state, tribal government, county, city, and neighborhood throughout the United States. In 2005, we began full implementation for the survey, and in 2006 we incorporated group quarters, such as nursing homes, college dormitories, and jails, into the survey—fulfilling our commitment to replace the long form in 2010. To date, we have released data for areas with populations of 65,000 or more. This year, we will release the first detailed information for areas with populations of 20,000 or more.
In FY 2009, we are requesting an additional $11 million to restore the Methods Panel to the ACS. The Methods Panel is a vital part of the ACS because it is the vehicle for testing new content and data collection methods, which help us contain costs and improve the quality of ACS estimates. It is, quite simply, a small investment in the future of the ACS. In FY 2009 our testing will include two distinct components of research, testing and evaluations. The first covers projects aimed at reducing various sources of non-sampling error, with the goal of improving data quality and achieving survey efficiencies. The second includes pre-testing and field testing of new content proposed for the ACS, as well as revisions to the current questions.
Our plans for FY 2009 demonstrate our commitment to fulfilling our constitutional mandate. In FY 2009 major field operations will be well underway. For example, in FY 2009, we will conduct the Address Canvassing operation for nearly 134 million housing units across the Nation. In addition, we will begin to validate the list of about 86 thousand group quarters. In FY 2009, we expect to deploy 680 people for the partnership program – most of whom will be specialists working in the field. With similar staffing levels for Census 2000, we established approximately 140,000 formal partnerships, and our goals for the program are no less ambitious this time around.
The census is not merely a federal effort. We rely on the participation and cooperation of literally thousands of communities throughout the United States. Reaching the residents of these communities, especially the hard-to-count, is the ultimate goal of the census and the fulfillment of the constitutional obligation. Our partners—advisory committees, national organizations, the faith-based community, elected officials, community and neighborhood leaders, and even the go-to person at the corner shop—are integral to this effort. The Census Bureau is planning an integrated communication, promotional, and marketing effort, incorporating partnerships and advertising, to reach every community. We have a contractor, Draftfcb of New York, working on an integrated communications plan, and we will brief our stakeholders on that plan starting later this month.
As part of our plans for the short-form only census, we have incorporated significant improvements over past censuses, notably automation and infrastructure. These efforts include the 2010 Decennial Response Integration System (DRIS) and the Field Data Collection Automation (FDCA) system mentioned earlier. The purpose of the DRIS contract, which was awarded in 2005 to the Lockheed Martin Corporation, is to ensure accurate and protected collection and storage of census responses, whether reported by paper form, hand-held computer, or telephone. It is on schedule and on budget.
It is also important to consider that other catalog of our nation’s wealth and prosperity, the Economic Census. The Economic Census, along with the Census of Governments, is conducted every five years, for years ending in ‘2 and ‘7. Last fall, approximately five million report forms were mailed to business establishments with employees. The due date for these forms was February 12, 2008, and collection extends into late summer 2008. The first results from the 2007 Economic Census will be released in March 2009. In FY 2009 we will complete activities for the 2007 Census of Governments. The Census Bureau’s FY 2009 budget requests $119 million for the Economic Census and $9 million for the Census of Governments.
The Economic Census provides the Nation with the most comprehensive, detailed, and authoritative facts about the structure of the U.S. economy, providing detailed information on 26 million business locations, covering 1,000 different industries. The 2007 Economic Census will cover 84 percent of economic activity. The Census of Governments includes information on 90,000 state and local governments and covers an additional 12 percent of GDP.
The Economic Census provides key source data for the Bureau of Economic Analysis’ National Income and Product Accounts, the framework for developing GDP estimates. It also provides weights and benchmarks for the Federal Reserve Board’s Index of Industrial Production and the Bureau of Labor Statistics’ Producer Price Index and productivity statistics programs. Moreover, the Economic Census updates the Census Bureau’s Business Register, the sampling frame for the Census Bureau’s current economic surveys, as well as providing benchmark estimates for the monthly principal economic indicators and annual surveys covering key sectors of the economy. Economic Census results are also widely relied on by businesses, trade associations, researchers, and academia.
The Economic Census and the Census of Governments provide the foundation for most of the nation’s economic statistics. They are, like the decennial census, a tremendous source of information. Together, the decennial and economic censuses offer compelling descriptions of the source of the nation’s wealth and prosperity – our economy and our people.
Just as the censuses offer broad and comprehensive assessments of the nation’s people and the economy, we also have other opportunities to explore information that can provide more detail about growing sectors of the economy and vulnerable populations.
One of the Census Bureau’s important goals for FY 2009 is the expansion and improved measurement of the services sector.
The FY 2009 budget request again includes an $8.1 million initiative to provide quarterly and annual coverage of all twelve service sectors. The service economy accounts for 55 percent of national economic activity, but currently our quarterly surveys cover only 17 percent and our annual surveys cover only 30 percent of GDP. When this multi-year expansion is completed, both quarterly and annual coverage of service industries will match the coverage of the Economic Census conducted every five years. The Economic Census, of course, will still provide a more comprehensive and detailed assessment than the quarterly and annual estimates.
The FY 2009 request also includes nearly $46 million for the Survey of Income and Program Participation – or SIPP. The FY 2008 Appropriation provided a total of $24 million to restore the SIPP, which is the major source of information on the economic well-being of Americans over time. It provides policy makers with demographic, income, and labor characteristics in order to measure the effectiveness of existing federal, state, and local programs. We will use that funding to reestablish the SIPP this September with the implementation of a new 2008 SIPP panel. Due to funding limitations there will be a 4-month data gap between the end of the 2004 Panel and the beginning of the 2008 Panel. Data gaps are not unprecedented, as we had one in the 2000-2001 time period. When the new panel begins in September, the SIPP will return to a traditional sample of 45,000 households. The FY 2009 request funds the SIPP at the full sample size. It also funds improvements to the SIPP program, such as updating the automated survey collection instrument, training enumerators in the use of that instrument, a reengineered processing system, an Event History Calendar Field Test, and security enhancements to laptops used in the field.
Finally, we have requested an additional $1.5 million to maintain and improve the quality of the Current Population Survey – or CPS. The CPS is a joint program with the Bureau of Labor Statistics (BLS) and provides detailed labor force characteristics of the civilian non-institutionalized population, the monthly unemployment rate, and official poverty estimates. The BLS is submitting a parallel initiative.
Mr. Chairman, I thank you for the opportunity to be here today and would be happy to answer any questions.
