Mr. Chairman, Mr. Clay, and Members of the Subcommittee:
It is a pleasure to appear before you again and to testify at this second hearing on the American Community Survey. The Subcommittee's leadership, Mr. Chairman, in providing a public forum for discussing the ACS is greatly appreciated.
The American Community Survey is Part of the Strategy for a Re-engineered 2010 Census
The ACS is one of three key components of our strategy for re-engineering the 2010 Census. As Secretary Evans testified in a Senate hearing on March 28: "While Census 2000 was an operational success and produced data of high quality, the process was costly, many people felt burdened by having to answer the long form questions, and the census was constantly at risk due to insufficient early planning and development, and disagreement on the design." If the Census Bureau has adequate resources early to build upon the successes of Census 2000 and to take advantage of lessons learned from Census 2000, then we can reduce operational risks for the 2010 Census, explore ways to further reduce the undercount and improve accuracy, provide more relevant and timely data throughout the decade, and contain costs.
While our strategic plan for the 2010 Census is still under development, we have identified what we believe are the three main components: 1) improving the accuracy of our geographic database and Master Address File, 2) eliminating the long form from the 2010 Census by collecting those data in the American Community Survey, and 3) re-engineering the census process through early planning.
We are planning improved geographic systems to ensure that there is a complete and unduplicated address list, that will facilitate automation and electronic data collection, and that will also provide the potential for significant efficiencies. We plan to investigate the use of Global Positioning System technology and satellite mapping imagery to link addresses to physical locations, which will allow accurate updating and improvement of Census 2000 maps and address information.
The ACS will simplify the 2010 Census requirements and allow the Census Bureau to focus exclusively on the basic count of the population. It will provide more current and more frequent detailed data for businesses and local, state, and federal policymakers for small geographic areas and allow the federal statistical system to keep pace with ever increasing demands for timely and relevant data. Data users now must still use 1990 long form data, data that will be 12 years old next year, when we begin the release of Census 2000 long form data. The ACS will allow federal policymakers, businesses, and state, local and tribal governments to make decisions using more current data and will improve the distribution of federal funds.
Early 2010 planning and development are necessary for a re-engineered process for the 2010 Census that takes advantage of the potential opportunities provided by having improved geographic systems and the ACS.
Mr. Chairman, in your opening remarks at last July's hearing on the ACS you said: "Today we are here to begin the process of eliminating the problematic census long form." The process of eliminating the census long form is well underway. That is why the ACS is one of the three key components of the plan for re-engineering the 2010 Census. It will allow us to eliminate the long form from the 2010 Census and collect data on congressionally mandated topics in an annual sample survey.
Census 2000 required the processing of more than 1.5 billion pieces of paper. About 60 percent of this paper was associated with the long form. It took a lot of talent, time, money, and effort to process that much paper. Implementing the vision, Mr. Chairman, of simplifying the decennial census by eliminating the long form will introduce opportunities for cost savings and improved coverage, and greatly enhance the Census Bureau's ability to streamline and modernize the census-taking process.
Replacing the long form with the American Community Survey will permit the Census Bureau to focus exclusively on conducting the basic count in 2010. It will allow us to streamline data capture, accommodate more multiple language response, increase mail response, and lower field costs. It will lead to better coverage of addresses in the census because it will require a continuously updated address list. Ongoing interaction with local officials will allow us to avoid the just-in-time address updating operations for the 2000 Census. The ACS will also provide current data to permit the Census Bureau to more effectively target areas where special procedures will be required in 2010. We are also looking at ways to use the highly trained ACS field staff as a cadre of key supervisory staff for the 2010 Census. Because they will have been collecting ACS data successfully in hard-to-enumerate areas, they will have developed relationships with local leaders that will prove invaluable in more completely enumerating the population in the 2010 Census.
Mr. Chairman, you asked that I address the issue of costs. The Census Bureau recognizes that there are substantial up-front costs in the early years of the decade for this strategy. For example, our preliminary estimate of the cost of the ACS for FY 2003, the first year of collecting data in every part of the country, is $131 million, subject, of course, to the regular budget process. As I stated in my letter to you, Mr. Chairman, of April 19, 2001, there will be only 9 months of nonresponse followup field work in the first year, which lowers the cost somewhat compared with all future years. Our initial estimates of life-cycle costs will demonstrate cost neutrality when we compare the estimated cost of repeating Census 2000 to the estimated cost of a re-engineered 2010 Census, which includes the ACS, geographic system modernization, and early planning. To achieve cost neutrality, and even have some potential for cost savings, while also providing so much more data throughout the decade, is a notable achievement.
A Balanced Design for the American Community Survey
Now, Mr. Chairman, I will turn to the ACS, and first describe how we worked with the statistical system and data users to balance issues related to cost, feasibility, data quality, and respondent burden in designing the program. The ACS is not designed to provide a headcount. That is what the decennial census is for. The Census Bureau has a population estimates program to provide updated population totals for years following the most recent census. The ACS will serve a similar role in providing updated population and housing characteristics.
Our goal in designing the ACS was to produce data comparable in quality to the decennial census long form for the smallest areas such as census tracts. One decision was how many years should go into the moving averages that will replace the long form estimates. We decided on a 5-year average for the ACS that will give more timely data throughout the entire decade, and will give much better information about change over time than a once-a-decade measurement.
Another decision was to determine how much sample was needed each year so that the 5-year averages would have a sample size to provide data of sufficient quality. We chose a sample size of 3 million because that will meet our goal of producing data based on 5-year averages that are comparable to census long form data for the smallest areas such as census tracts. These data will be available in 2008, 4 years earlier than from a 2010 long form, and they will be updated annually thereafter. Moreover, the first data for communities of 65,000 people or more will be available in 2004, about 6 months after the 2003 data collection year ends, and every year thereafter. The first data for communities with between 20,000 and 65,000 people will be available in 2006, and also updated annually after they are first available.
The Census Bureau will mail questionnaires to each address in the sample and will use purchased lists of telephone numbers to interview households that do not mail back their form. We will select a one-in-three sample of the addresses not interviewed by mail or telephone and conduct personal visit interviews. Since personal visit interviews are very expensive, this sub-sampling will reduce costs while providing the desired level of accuracy. The quality of data is improved by using well-trained, permanent interviewers with computers to complete interviews. Computer-assisted interviewing by telephone or in person allows the Census Bureau to incorporate consistency checks of the data into the collection process. Data quality is further improved by the use of a permanent staff to code write-in entries for processing.
The American Community Survey will be based on making the most efficient use of existing systems and infrastructure. It will not require additional offices, but will be run out of our current regional offices. Field interviewers, who work on other surveys in addition to the ACS, will use existing laptop computers, and laptops from Census 2000 can be used as replacements, to the extent possible. As I noted in my letter of April 19, for FY 2003 the ACS will require about 1,230 full-time staff and 3,800 part-time staff. We will not require special data capture centers, as that will be accomplished in our processing center in Jeffersonville, Indiana. The Census Bureau is working to provide electronic versions of maps that will reside on the interviewers' laptops; this will reduce the massive requirements for paper copies of maps we encountered in the decennial census. We do not anticipate, at this time, any advertising expenditures for the ACS because its small monthly sample size would make it inefficient to use mass marketing methods such as we needed for the decennial census. We do expect to use community partnerships, which we have found to be very effective in the early test sites.
Mr. Chairman, you also asked whether the ACS would replace or incorporate any additional Census Bureau surveys. While there are no current plans to replace any existing surveys, the ACS can help make the entire federal statistical system more efficient in a number of ways by improving existing programs. They include the Current Population Survey, the National Crime Victimization Survey, and the Consumer Expenditure Survey. The Bureau of Labor Statistics will use ACS data as part of its Local Area Unemployment Statistics program to make improved monthly estimates of unemployment. And the Census Bureau will use the ACS data, along with the Current Population Survey data and administrative records, to improve poverty estimates for school districts.
Data for Small Areas
The fact that the ACS sample size and design will not provide data for the smallest areas until 2008 has led some to raise the concern that the ACS may be treating rural areas and urban census tracts unfairly. The Census Bureau takes this concern very seriously. Indeed, we wish it were possible to begin providing small areas with high quality, current data right away. But that would basically require replicating the decennial long form every year, and that is not acceptable in terms of costs or burden on respondents. Once again, we are faced with balancing competing demands.
But the Census Bureau takes even the perception that small areas are being treated unfairly very seriously and has worked with data experts to allay those concerns. For example, the Census Bureau contracted with WESTAT to bring together rural experts and get their suggestions for improving the ACS from the perspective of rural areas. The group of a dozen experts, who met in May 1998, expressed strong support for the ACS and we are pursuing their suggestions for further research and changes that might make the ACS even more valuable to rural areas.
At last year's hearing, you heard from Congresswoman Jo Ann Emerson; Ms. Barbara Welty, representing the National Association of Towns and Townships; and Mr. Chuck Fluharty, Director of the Rural Policy Research Institute, on this issue and about the importance of having current data to support rural policy initiatives. If I may quote from Mr. Fluharty's statement last year:
"While some statistical limitations for rural areas within the American Community Survey may exist, the more timely, cost-effective multi-year estimates produced by ACS are far superior to existing data produced by the decennial Census."
And this is the key point--even the smallest areas will have data 4 years earlier than they would if we had no ACS and included a long form in the 2010 Census.
So while concerns have been raised about data for small areas, the Census Bureau is confident that the current ACS design yields a major improvement over the existing situation. We need to understand that the glass is more than half full and to fill it all the way would require unacceptable trade-offs in terms of costs and respondent burden.
Data for Small Population Groups
We have designed the American Community Survey to provide the same quality data as Census 2000 for all groups, regardless of size, and we plan to monitor the survey to ensure that this is the case. As noted earlier, the ACS does not count the population; it estimates their characteristics. To get accurate measurements, we need high response rates from all groups.
We have devoted considerable time to discussing the question of data for small population groups with our Race and Ethnic Advisory Committees. We are evaluating their suggestions, together with our experiences collecting data from these groups in Census 2000.
Working closely with them, we will focus on techniques and strategies to ensure that small population groups participate in the survey. We are exploring using language assistance guides similar to those used in Census 2000; revising the mailing package to elicit a better mail response rate; and placing public service announcements in newspapers and electronic media serving minority populations.
The permanent staff of field representatives will establish ongoing relationships with the communities they are working in, thereby enhancing trust and willingness to participate.
Timely and Relevant Data to Manage Federal and Local Programs
We believe that the American Community Survey is the cornerstone of the government's effort to keep pace with the country's ever-increasing demands for timely and relevant household data. If we are to continue leadership in providing information to the knowledge society, we must invest in full development of the ACS. In our web-based society of the 21st century, it is no longer acceptable that planners and policy-makers must use data that are up to 12 years out of date.
Eliminating the long form from the once-a-decade census will give us an incredible head-start toward having a successful 2010 Census at a reasonable cost. Over the years, the long form included in the decennial census has met the Nation's need for detailed data essential for governing, by providing reliable, consistent data for small geographic areas once a decade. To eliminate the long form without an alternative method of collecting detailed population and housing characteristics would leave the Nation with a knowledge void at a time when timely and accurate information is required to design and manage federal programs and distribute federal funds fairly.
Without an appropriate alternative to the decennial census long form--the ACS--Congress would not have the data to run the programs it has written into law. It would not have a rudder to guide the ship of state.
The ACS will provide timely data to help the Congress, federal program managers, and the "end users" in every community in America. It recognizes the continuing roles of the Congress and federal agencies, while also elevating the importance of partnerships with other levels of government and those directly affected by the decisions based on data. Data collected by the ACS will help Congress evaluate and modify federal programs and will provide up-to-date information for congressional districts and states, as well as smaller areas, enabling services to be targeted to maximize the impact of available resources at all levels of government. The ACS will provide a critical new source of data that will allow the Congress to evaluate programs below the state level and determine accountability. When fully implemented, it will collect data in every county, American Indian reservation, Alaskan Native area, and Hawaiian homeland, as well as Puerto Rico. Because data from the ACS will be available every year, Congress can determine change over time and measure the results of federal programs. It means greatly improved annual estimates of population and housing characteristics and how they change over time. The up-to-date estimates from the ACS will benefit, for example, welfare reform, funding for educationally disadvantaged children, and programs for the elderly.
The welfare reform legislation provides performance bonuses to states. In some cases, no data are currently available to measure the results of state programs. The ACS will collect the information about income and family structure that is needed to measure poverty and eligibility for benefits, as well as age, marital status, the presence of children, and educational attainment and school enrollment. In addition, it will shed light on the results of state and local welfare-to-work initiatives.
The American Community Survey dramatically improves poverty estimates for groups such as children and the elderly, and provides current information on changes in family structure. The Department of Education currently uses the estimates of poverty for children from the Census Bureau's Small Area Income and Poverty Estimates program in allocating annually over $8 billion in Title I funds to school districts. The ACS will dramatically improve these estimates for school districts. The "Leave No Child Behind Act of 2001," currently under consideration in the Congress, specifically mentions the ACS as the source of data on the number of children with limited English proficiency--data to be used in administering grants to the States.
The ACS provides information about all age groups, including the current elderly population and the characteristics of people nearing retirement age. This information is essential for projections about the needs and resources of people who are eligible to enter the retirement system and can help target screening programs for particular health conditions.
While I have focused on federal uses of the data today, the American Community Survey is providing current data from 21 of its 31 test sites to address real-life issues in rural and urban communities, and I think it's important to mention just a few illustrations of local uses.
A sparsely populated, rural county--Fulton County, Pennsylvania--is using the ACS data to develop an employment and training services system, a rural transportation system, and a "Help Line" for families to address health care and child care needs.
Bronx County, New York, will use ACS data to identify and develop intervention strategies for juvenile diabetes in special populations.
In Springfield, Massachusetts, community leaders and public safety officials are using ACS data to develop a youth violence prevention program for teens.
These are just a few examples of how having up-to-date information, instead of information that is 12 years old, will allow communities to provide services more efficiently. ACS data are also being used to help localities meet federal program requirements.
Multnomah County, Oregon, will use ACS data as a source for enrollment forecasts that are used to justify support under various school programs. In addition, Multnomah County used 1996 ACS data for a study of students who were in poverty, a critical input to several federal programs. And ACS data on the number and demographic characteristics of children and families receiving services are being used by the county to develop Head Start Community Assessment reports.
Next month, the we plan to issue additional data from the 21 test sites and national and state-level data from the nationwide operational test that collected data from a national sample of about 700,000 households. We conducted this test--the Census 2000 Supplementary Survey--to ensure that the ACS can be conducted separately from, but during the same period as, a once-a-decade census.
Developing American Community Survey Content
In conjunction with the Office of Management and Budget, we have established a jointly chaired Interagency Committee charged with balancing respondent burden with the legitimate information needs of the Congress and the federal government. The Interagency Committee is working on reviewing the content of the ACS in a process similar to what we've done for the decennial census long form. OMB has asked relevant federal departments and agencies to document legal requirements for data, the level of geography needed, and for what population groups. This information is expected to be complete by the end of August.
For 2003, we anticipate the questionnaire will be similar to the current ACS questionnaire. Since the survey is designed to produce 5-year estimates, we believe the content should remain reasonably consistent over time. We would plan to submit to the Congress in 2006 the revised content of the 2008 questionnaires, following the decennial model. In the interim, and through the Interagency Committee, all questions will be reviewed, those no longer required will be dropped, others can be re-worded and tested, and congressionally mandated topics can be added.
There was considerable discussion of the long form last year as an "invasion of privacy." This was partly because the long form was administered to about 20 million housing units all at one time. The American Community Survey provides a different atmosphere and will reduce public concern about intrusiveness.
Instead of about 20 million housing units receiving the census long form all at one time, in each monthly survey, only 250,000 housing units will receive the ACS questionnaire at any one time. This is 1 in every 480 households in any given month. In any given year, only 1 in 40 households, or about 2.5 percent of the total, will receive the ACS questionnaire.
Instead of having a large army of temporary enumerators, the ACS field interviewers will be highly trained permanent staff who will be better prepared to deal with the public's questions about the form. Our experience thus far with the ACS has been that our trained interviewers have achieved good cooperation with few complaints from the public. They can explain that the confidentiality of the data is strongly protected, they are used only for statistical purposes--not for regulation or law enforcement--and each questionnaire item is linked to a federal program.
Having said that, the Census Bureau takes questions and concerns about intrusiveness and privacy seriously. We are aware of the time pressures confronting people and of the concerns we all have about privacy and confidentiality. We also recognize and respect those who have philosophical objections to government activities, government programs, and government data collection. In fact, the Census Bureau has a 60-year history--going back to the 1940 census--of working to reduce the number of questions and the number of households that would have to answer the longer set of questions. Before 1940, everyone was asked the "long form" questions; that year, some of the questions were asked of only one-in-four people. For the 1960 Census, the percentage of households that received the long form was still one-in-four, but this was reduced to one-in-six by the 1980 Census. And in Census 2000, the short form was reduced to seven items--the shortest in 180 years. Weighed against the ever increasing demands for new questions, including requests from the Congress, this is evidence of the Census Bureau's sensitivity to this issue.
Conclusion
In summary, Mr. Chairman, as part of the strategy to re-engineer the 2010 Census, the American Community Survey will improve the way we take the census by eliminating the long form, simplifying the 2010 Census requirements, and allowing the Census Bureau to focus exclusively on the basic count. It will provide more frequent detailed data for all geographic areas regardless of size, so that the Congress and federal agencies will have up-to-date information to administer and evaluate programs. And it will contribute to a more efficient statistical system and allow us to keep pace with ever increasing demands for timely and relevant data.
That concludes my testimony and I will now be happy to answer any questions.