PREPARED STATEMENT OF

KENNETH PREWITT

BUREAU OF THE CENSUS

UNITED STATES DEPARTMENT OF COMMERCE

Before the Subcommittee on

Commerce, Justice, State,

the Judiciary, and Related Agencies

Committee on Appropriations

House of Representatives

April 6, 2000

Mr. Chairman and members of the Subcommittee:

Thank you for the opportunity to discuss with you today the Census Bureau's Fiscal Year 2001 budget request.

Mr. Chairman, for FY 2001 the Census Bureau is seeking funds to fulfill its constitutional and legal obligations to issue the results of Census 2000. We are also seeking funds to modernize the Census Bureau so that it will continue to play a leading role in establishing the information infrastructure of our knowledge society. Let me talk briefly about each of these points and then provide greater detail about our budget request.

Although we will complete data collection activities for Census 2000 this fiscal year, the census will not be over. Through FY 2000, we will have invested $6.2 billion in planning and conducting Census 2000. The payoff for our investment will come when we process, tabulate, and disseminate quality and timely data for apportionment, redistricting, the allocation of Federal funds, re-benchmarking household surveys, community planning, and the myriad of other important purposes for which we conducted this census. And that will begin in FY 2001. By the end of this calendar year, we will have built an enormous storehouse of information, bulging with billions of bytes of data about our people and our housing. The funding we are requesting for the decennial census in FY 2001 will allow us to organize and quality-check this valuable data, and make sure it flows where it is needed.

While much critical work for Census 2000 remains to be done, the diminishing level of Census 2000 activities in FY 2001 will allow the Census Bureau also to focus on modernization projects that will position us to collect, process, and deliver timely, relevant and cost-effective data about the economy and the American people through other Bureau statistical programs. Economic growth is high and the rate of social change is increasing. So we are requesting new funding to position ourselves to be responsive to the growing demands from policy-makers for more timely and relevant demographic and economic information on important new or rapidly changing aspects of the Nation's people and economy.

Canals, shipping lanes, railroads, and highways have been the infrastructure of our industrial economy. Today, in the era of the Information Superhighway, information is an essential part of the infrastructure of our knowledge society. In determining how much we as a society should invest to improve that infrastructure, we must realize that census data are a "public good," critical to the functioning of our economy. As a "public good," census data are used for important constitutional and political purposes, in economic measurement and planning, funds allocation, and in community planning. These are the governmental uses of our data, but the data also contribute widely to the overall well-being of our economy through a variety of private sector uses--often after the data are repackaged. While the data provide broad private sector benefits, only the government can organize and support the collection of these data, ensure their quality, and lend them credibility. That is why the Census Bureau is seeking to meet its larger public responsibilities by requesting funding to begin the process of modernization. Without a strong, vibrant, and modern Census Bureau our Nation will not have a sound infrastructure for the 21^st century.

In this regard, the American Community Survey (ACS), part of our Continuous Measurement program, is our single most important improvement in Federal household statistics since the middle of the last century and I believe it is the cornerstone of our effort to keep pace with ever increasing demands for timely and relevant data. If the Census Bureau is going to remain a major contributor in the knowledge society, we must invest in full development of the ACS. In our web-based society of the 21^st century, where information is instantly available at everyone's fingertips, it is no longer acceptable that planners and policy-makers must wait 10 years for the data they need to do their work. Not only will the ACS provide more frequent detailed data for small geographic areas but it will also revolutionize the way we take the decennial census

There are other elements of our modernization effort that I will discuss in more detail later, but let me just mention some of them now.

New, updated, coordinated samples for the major recurring household surveys that will improve the accuracy of the surveys without increasing their costs.

Improving the way we conduct business by taking advantage of modern, efficient technologies.

Measurement improvements for electronic commerce, minority-owned businesses, and U.S. exports.

Improved measures of economic well-being from the Current Population Survey and the Survey of Income and Program Participation.

These are not glitzy initiatives, but they are the nuts and bolts components of an improved information infrastructure.

Now, let me speak about infrastructure in the older sense of the word. The Census Bureau's talented and dedicated staff are the key to all our efforts to keep pace with society's demand for data. They deserve a modernized and safe environment to work in. I am referring to the nearly 60 year old, shabby Suitland facility where most of our headquarters staff now work. Leaky roofs, bursting pipes, asbestos, a contaminated water supply--these are just some of the health and safety concerns that Census Bureau employees face daily. We are seeking funding to cover some of the internal planning effort essential for clearly and fully identifying our space and technology needs. The General Services Administration (GSA) will need this input in order to complete its full renovation or reconstruction of our facilities.

The Bureau is requesting direct appropriations of $719,205,000 for domestic discretionary spending; of that, $173,826,000 is for current statistical programs, and $545,379,000 is for periodic census programs as presented in Exhibit A. Now, I will discuss in more detail some of the specific aspects of our FY 2001 request.

As we enter fiscal year 2001, we will be winding down the huge, labor-intensive field data collection operations for Census 2000. The vast majority of enumeration activities, including nonresponse followup, will have been completed and hundreds of thousands of temporary employees will be released by the Bureau. As in past censuses, the downsizing of the workforce will require significant effort for a smooth transition to post-Decennial Census staffing levels. The 520 Local Census Offices will complete data collection activities, including the Accuracy and Coverage Evaluation and will enter the final stages of close out. Data capture activities will be nearing completion, and we plan to close the three contracted Data Capture Centers by the end of the first quarter of FY 2001. We will continue with the clerical coding of long form, write-in data until the fourth quarter at the National Processing Center in Jeffersonville, Indiana, the permanent Bureau data processing facility.

While the budgetary needs of the Decennial Census decrease significantly, it is vital that full funding be provided in FY 2001 to enable the Bureau to close down the field operations in an orderly manner. Although the most costly Census 2000 activities will have been completed, Census 2000 will not be over. We still need to process, tabulate, and disseminate data on time, which requires completing complex coding, editing, and other processing steps. The Census 2000 results for apportionment, redistricting, and allocation of Federal funds will be presented on a flow basis beginning with delivery to the President by December 31, 2000 of the state level population counts, necessary to apportion congressional seats. The data below the state level for use in legislative redistricting will be complete by March 31, 2001.

In FY 2001 we must also archive the data and complete extensive evaluations and operational assessments that will be used to document Census 2000 operations and systems performance, and which will provide information critical to early planning for the next decennial census. We will also respond to concerns from State, local, and tribal governments, as well as community-based organizations, regarding the census counts.

In 2001, we will have the opportunity to continue the program to modernize how we collect the basic socioeconomic data which has been collected on the long form of the decennial census since 1940. The American Community Survey or ACS (described more fully under "Continuous Measurement") and the census long form transitional database will allow us to eliminate the long form from the 2010 census and to plan a census focused exclusively on population counts.

But this can happen only if we continue the work to ensure that the ACS is an acceptable replacement for the census long form and if we continue the working partnership between the Congress and the Census Bureau - working closely together to make sure that our modernization efforts are meeting government needs.

In FY 2001, the Census Bureau is requesting $392,798,000 in new budget authority for the Decennial Census.

Following each decennial census the Bureau also undertakes additional work to improve Federal statistics. This additional work is conducted in two ongoing, cyclical programs: the Demographic Surveys Sample Redesign program, and the Intercensal Demographic Estimates program.

The goal of the Demographic Surveys Sample Redesign program is to draw new, updated samples for the major recurring household surveys, thereby maintaining the accuracy of, and confidence in the major federal socioeconomic indicators. The Census Bureau is requesting a program increase of $3,688,000 in FY 2001 to fund the activities of survey redesign that are common to all household surveys conducted by the Bureau, and use the decennial census as a baseline. The samples drawn from the 1990 census will be exhausted by 2004. Therefore, it is imperative that we select new samples for the surveys for the next decade based on Census 2000 information. Without new sample frames, we will have severely distorted estimates of the Nation's unemployment rate, generate poverty statistics, or create key components of the consumer price index. The long term impact of ignoring the cyclical funding request for this program would have a catastrophic effect on all of these surveys. The Census Bureau request for Demographic Surveys Sample Redesign is $8,457,000. The Census Bureau bears the core costs of this effort; work linked to specific surveys is charged to sponsoring agencies.

INTERCENSAL DEMOGRAPHIC ESTIMATES

The Intercensal Demographic Estimate program develops updated population estimates in years between decennial censuses for states, counties, metropolitan areas, and urban places. These post-censal estimates are needed because the births, deaths, and migration that occur between censuses change the size and distribution of the population. These estimates are released periodically during the decade, and there are numerous Federal and State programs that base funding on them. We are requesting $5,583,000 to continue the Intercensal Demographic Estimates program at the current level.

The Continuous Measurement program includes the American Community Survey (ACS) , which is already a success story. In FY 2001 we will continue to collect data in 31 sites across the nation to allow intensive analysis for small geographic areas, such as census tracts, that require accumulations of several years of data. When fully operational in 2003, the ACS will collect current, small-area census long form type data for all areas of the country. By gathering these data on a more frequent basis and forming partnerships with other Federal agencies and State and local officials, the Census Bureau is able to enhance the value of this information tool. For example, the Census Bureau will work with the Department of Justice in a crime mapping and data-driven management initiative. The community policing perspective of this initiative requires current, small-area data not only on crime but also on factors which create crime risks and protect against those risks. American Community Survey data on clusters of poor housing, school dropouts who are not employed, and people who walk to work tell local officials about crime risk factors.

A program increase of $3,385,000, for a total of $25 million, will support critical tests for conducting the American Community Survey in unique geographical areas and with unique populations groups, including the methodology for interviewing in rural Alaska and the feasibility of conducting a mail survey in Puerto Rico.

While many people associate the Census Bureau only with the decennial census or with population estimates and projections, there is a vast array of other familiar information the Bureau collects and provides to policymakers on the socioeconomic, demographic, and housing characteristics of the Nation's people. Two primary surveys are the Current Population Survey, or the CPS, and the Survey of Income and Program Participation, often referred to as the SIPP.

The CPS is the primary source of information on the labor force status of the population, and is funded as a joint program with the Bureau of Labor Statistics (BLS). It measures how much of the labor force is employed and unemployed, and refines employment data by various other characteristics such as age, sex, ethnic origin, and veteran status. In addition, the March Annual Demographic Supplement to the CPS provides the official estimates of annual income, poverty, and health insurance coverage, while other supplements address topics such as voting behavior.

The SIPP provides detailed data on changes in income, poverty, health insurance coverage, program participation and eligibility, disability, non-cash benefits, and general demographic characteristics, by interviewing a representative sample of people for three years. The SIPP estimates of economic well-being are more comprehensive than those from the CPS, including information on benefits received from government programs and how they affect economic well-being over time.

In response to a National Academy of Science recommendation, we are proposing a $6,100,000 program initiative for Improved Measurement of Economic Well-Being. It will provide the basic data needed to improve the 35-year-old poverty definition based on cash income only as well as to provide new indicators of other important socioeconomic conditions by transforming the Survey of Income and Program Participation (SIPP). The new design of SIPP will have three overlapping three-year panels that will permit accurate, consistent, year-to-year comparisons. In other words, SIPP will provide reliable estimates of both current levels and year-to-year changes in such important economic indicators as the employment rates of the disabled population and variations in their employment experience, median household income and poverty rates, the use of public versus private child care services, rates of participation in and eligibility for transfer programs and multiple program participation, health insurance coverage rates of the newly unemployed and of children in such families, and asset ownership by population subgroups. By increasing the number of households interviewed three times a year from roughly 36,000 now to 47,000 in 2001 and 61,000 in 2002, the initiative will allow the SIPP to provide clear, reliable, annual "snapshots" of the current economic status of the Nation's population as a whole.

Furthermore, in the FY 2000 appropriations process, because the base funding for the Current Demographic Statistics programs was reduced by over $5,000,000, the Survey of Income and Program Participation (SIPP) experienced significant losses. For example, we canceled the last wave of the 1996 SIPP Panel; canceled use of monetary incentives for respondents to improve response in the 2000 SIPP Panel; and postponed work on the planned conversion of the automated questionnaire and control systems from DOS-based software now in use to a graphical user interface (GUI) based software. Because of the dropped monetary incentives, we expect a substantially higher non-participation rate for this survey, resulting in a reduction in the quality and reliability of the data. Delaying the conversion to GUI places the Census Bureau substantially behind other government and private statistical organizations in adopting a productivity enhancing software standard. We must recover from these losses.

Our FY 2001 budget includes $3,721,000 for the Current Demographic Statistics programs to restore a portion of the cuts. We will use approximately $1,700,000 to recontact and interview non-respondents to improve SIPP data quality, and approximately $2,000,000 for GUI conversion to enhance data processing productivity.

For our Current Demographic Statistics programs, the Census Bureau is requesting new budget authority of $58,299,000.

More than 100 annual, quarterly, and monthly surveys carry forward key national economic statistics on a current basis -- including business, construction, manufacturing, Federal expenditures, and general economic statistics. The Census Bureau is requesting $111,628,000 in new budget authority for Current Economic Statistics programs for FY 2001.

The Census Bureau is responding to the requests of policymakers for new, better, and more timely data by proposing three new policy initiatives in our Current Economic Statistics programs. We are working closely with the Bureau of Economic Analysis to provide the most complete and up-to-date measures of the economy possible.

• According to a 1999 Department of Commerce report, the Internet may be used for more than $1 trillion worth of transactions between businesses in just a few years. The Measuring Electronic Business initiative requests $8,500,000 to establish an on-going program to measure e-commerce and to fund development and implementation in our business programs of electronic data collection methods, a step that holds the promise of significantly reducing reporting burden yet improving data accuracy at the same time. The Bureau of Economic Analysis has a complementary $3 million E-Commerce measurement initiative in its FY 2001 request.

• Currently, statistics on minority-owned businesses are collected every five years, in conjunction with the Economic Census. The Survey of Minority Owned Business Enterprises initiative (SMOBE) requests $1,500,000 for an annual survey to respond to the need for more timely data by government and business customers who use the data to make informed policy, program, and business decisions.
• Studies by the Census Bureau suggest that our official estimates of U.S. exports are undervalued by anywhere from three to seven percent. Changes in trade practices such as the rapid growth of the air courier industry, erroneous filings of documents, and lax compliance with reporting requirements account for much of the gap. We are requesting $1,400,000 for the Improvement of Export Coverage to support a multi-pronged effort to remedy the problem. Targeted outreach and education efforts will help ensure that exporters are aware of their responsibilities under law and help them report exports correctly. The initiative will also develop an assessment program to monitor the accuracy of data and deter willful noncompliance and launch a program directed at improving the quality of estimates of low valued exports, those under $2,500.

Furthermore, a reduction in base funding of over $11 million for the Current Economic Statistics programs in FY 2000 required us to curtail the Current Industrial Reports program; postpone implementation of the North American Industry Classification System (NAICS) in current economic surveys; eliminate the Shipments to the Federal Government Survey; abandon work to develop time series data for NAICS service industries; and scale back processing and program specific contract support. These temporary suspensions will have long-term effects. For example, the early conversion of Federal statistics to NAICS will be slowed down and made more limited. Because we lack funds to collect data on many of the newly classified service sectors, we will remain uninformed about a good portion of the new sectors the NAICS was designed to identify. These include, for example, business, computer, and trade schools; the administration of insurance and pension funds; diminished detail for the health care industries and the publishing industries; and a void in expense data for communication industries.