Subject: Census News Brief #10 Date: Wed, 27 Aug 2003 19:13:58 EDT National Research Council Panel Urges More Informed Planning for 2010 Census; Census Funding Clears House Of Reps Intact; House Subcommittee Hears Strong Support; New Cost Estimate, for ACS; and more. In its third report, issued last month, the National Research Council’s Panel on Research on Future Census Methods urged the Census Bureau to complete evaluations of Census 2000 quickly in order to assess more fully the costs and benefits of design options for 2010. The science panel also expressed concern about the direction of plans to update and improve the Master Address File (MAF), which represents the universe for household enumeration. The panel, chaired by Dr. Benjamin King, was established to review early planning for the 2010 count. Its second interim report (the panel previously published a letter report, as well) focuses primarily on modernization of the Census Bureau’s geographic databases and on development of a technical infrastructure for the 2010 census. The Census Bureau often describes its 2010 census planning strategy as a “three-legged stool” comprised of (1) MAF/TIGER Enhancements Program; (2) replacement of the census long form with an on-going American Community Survey; and (3) early integrated planning for a short form-only census. The Academy panel cited “early integrated planning as perhaps the one [component] most crucial to a successful 2010 census – and, in some respects, the hardest to accomplish.” It noted that the technical infrastructure for the census, needed to develop a logical design process for 2010, was “functional but developed at high cost and high risk, without adequate time for development and testing” for Census 2000. The panel said it was “strongly impressed” by early efforts to reengineer the design process and encouraged the bureau to develop comprehensive business process models for 2010 quickly, and to appoint a “system architect” for the census. (Technical infrastructure refers primarily to software programs and computer and telecommunications systems that support census operations.) The panel also evaluated plans to modernize the Census Bureau’s geographic database, called TIGER (Topologically Integrated Geographic Encoding Reference System), and comprehensive housing list (MAF), for the 2010 census. The report is generally supportive of the Bureau’s efforts to realign features in TIGER. The “plausibility” of the realignment timetable would be “bolstered considerably,” the panel suggested, by a detailed work plan that includes a schedule for updating county information, and by specifying a process for detecting geographic changes once the initial updating is completed. The report is far more critical of proposals to improve the Master Address File. The modernization program, the panel concluded, “falls seriously short” with respect to the MAF, and is lacking a comprehensive plan to add new addresses, identify duplicates, and generally ensure a complete and accurate list. “The Census Bureau’s current strategy shows signs of repeating costly errors from the 2000 experience,” the panel wrote, noting that despite using the Postal Service’s Delivery Sequence File and instituting a pre-census local review of address lists, the Bureau conducted a costly, 100-percent block canvass of addresses before Census 2000. The panel recommended that the Census Bureau clarify how it will improve the accuracy of the MAF, including how it will apply its experiences from the 2000 census; appoint a coordinator to oversee MAF development and maintenance; specify plans for local partnerships, with an emphasis on improved communication and incentives for local sharing of information; and strengthen the role of evaluations in developing plans to update the MAF for 2010. The report also contains initial feedback on the American Community Survey (ACS). The panel encouraged Congress to fully fund the ACS over the long term, while also noting statistical issues that required further research, including the weighting and estimation scheme and the relative accuracy of ACS and long form data. Overall, the panel concluded that the strategy for the 2010 census “needs to be more fully informed” by cost-benefit analyses of different approaches. “To date,” the panel wrote, “the plan for the 2010 census has been presented to the panel with little supporting analysis.” The National Research Council (NRC) is the principal operating agency of The National Academy of Sciences and The National Academy of Engineering. The census panel was constituted by the NRC’s Committee on National Statistics. The report, Planning the 2010 Census: Second Interim Report, is available on the National Academies Press website at http://www.nap.edu. The panel will complete its work and issue a final report at the end of the year. Census funding update: By a vote of 400 – 21, the U.S. House of Representatives passed a $41.2 billion spending measure that includes funding for the Census Bureau, before adjourning for its August break. The Fiscal Year 2004 (FY04) Commerce, Justice, and State, The Judiciary, and Related Agencies Appropriations bill (H.R. 2799) allocates the full $662 million for Census Bureau programs requested by President Bush. The White House said in a Statement of Administration Policy that it supported passage of the bill, and “appreciates the full funding of Census Bureau programs,” but that it had several concerns about specific provisions of H.R. 2799. The administration urged legislators to restore $10 million cut from the President’s request for the Commerce Department’s Economic and Statistics Administration, which houses the Bureau of Economic Analysis (BEA). Only one provision of the mammoth bill, which would block implementation of a controversial Federal Communications Commission rule on media ownership, drew a veto threat. The Senate delayed consideration of a Commerce spending bill until after the August congressional recess. Congress resumes legislative work on September 3. For details on Census Bureau and BEA funding in H.R. 2799, see the July 10 and July 21 Census News Briefs. House panel hears testimony in support of the ACS: The House Subcommittee on Technology, Information Policy, Intergovernmental Relations, and the Census held an oversight hearing in May to review plans for the proposed American Community Survey. It was the panel’s first hearing on Census Bureau programs since its creation as part of the Government Reform Committee early in the 108th Congress. Rep. William “Lacy” Clay (D-MO), the subcommittee’s top Democrat, is the only panel member with previous service on a census oversight subcommittee. In opening remarks, Chairman Adam Putnam (R-FL) noted the Census Bureau’s historical efforts to meet “the continually changing needs of our Nation for timely, quality data,” and said Congress must now consider “another significant evolution”: replacing the traditional census long form with the ACS. To win support for ACS funding in Congress, the chairman said, the agency must show that the ACS “will eliminate duplicative surveys at the Census Bureau,” thereby saving money. “I simply would find it unbelievable that no surveys could be eliminated” when the ACS is fully implemented, the congressman said. Chairman Putnam also pointed to privacy as another significant issue the Census Bureau must address. The challenge, he said, is “to obtain the information that is needed to make informed decisions while at the same time respecting the privacy rights of the public.” While noting that the Census Bureau benefits from “the most protective privacy law on the books,” the congressman challenged the agency “to explore new and innovative ways to solicit voluntary cooperation” from the public. The bureau is evaluating a test conducted earlier this year at the request of Congress, comparing the effects of mandatory versus voluntary response on the cost and quality of ACS data. Rep. Clay said he remains concerned about the quality of data the ACS will produce and about adequate funding, as well as the long-term cost of the ACS. He noted that the delayed start of the ACS would allow stakeholders to evaluate additional years of data from test sites and the national sample. The Census Bureau plans to release data tables from the 2002 Supplementary Survey and test sites this fall. The subcommittee heard testimony from the Under Secretary of Commerce for Economic Affairs, the Census Bureau director, and a panel of five witnesses representing a range of census data users. Under Secretary Kathleen Cooper said the ACS will “revolutionize” the way we take the census and meet the nation’s data needs. She described the survey as a “well-developed program” that is “at full implementation” in 31 test sites across the country and producing “consistently high quality data” from these sites and a national Supplementary Survey of 700,000 households. Dr. Cooper reminded panel members that the content of the ACS questionnaire is essentially the same as the census long form it is designed to replace, because the Census Bureau must still meet federal programmatic requirements for data. Director C. Louis Kincannon noted that about $200 billion a year in federal program funds is distributed on the basis of census data. Referring to the revised plan to launch the ACS nationwide in the final quarter of Fiscal Year 2004, Mr. Kincannon said tract level data from the survey would be published in 2010, two years before it would be available if the 2010 census included a traditional long form. The first characteristics data, for places of 65,000 or more persons, would be available starting in 2006, based on the first full year of data collected in 2005. Mr. Kincannon acknowledged that tract level estimates from the ACS would be subject to higher sampling error than similar census long form data, because the five-year aggregates used to produce ACS tract data involve a smaller overall sample (15 million households over five years) than the long form (roughly 20 million households). Sampling error, the director said, would be offset “by more complete responses [to the survey] that the permanent staff of field representatives will collect” during telephone and in-person follow-up interviews to unresponsive households. A re-engineered 2010 census that includes only a short form, with the ACS replacing the long form, and a modernized address and mapping system, will cost $11.2 billion, Mr. Kincannon told the subcommittee. A traditional census plan would cost at least $12 billion, he said. In response to a question from Chairman Putnam, the director said the annual cost of a fully implemented ACS could be $165 million. Previous estimates submitted to the census oversight subcommittee in April 2001 pegged the cost of the ACS at $131 million for the first year, which under the original implementation plan would have included only nine months of field follow-up visits, the most costly part of the survey. Five witnesses presented the views of data users in state and local government, business, and demographic research for a variety of applications. All expressed support for the ACS and offered examples of how annual estimates of social and economic characteristics would improve decision-making for the data users they represented. Thomas Reardon, Executive Director of the Fulton County Partnership, McConnellsburg, Pennsylvania, discussed how annual data from the ACS test site in his county helped the organization accurately estimate and distribute flu vaccines for their small rural community. “[T]he more information we have about the people we serve, the more efficient we can be with the resources we have to help them,” Mr. Reardon testified. Joseph Salvo, Director, Population Division, New York City Department of City Planning, called the ACS “a continuous barometer of the social and economic condition of the nation’s communities.” Noting that the September 11, 2001 terrorist attacks diminished the usefulness of Census 2000 long form data for broad sections of the city, Dr. Salvo emphasized the importance of current information for the cost-effective delivery of government services. He also said the ACS would collect more accurate information on population and neighborhood characteristics than the census long form, citing successful follow-up interviews with unresponsive households in the ACS Bronx test site, where a significant number of households failed to complete long forms in Census 2000. Joan Naymark, Director of Research and Planning for Target Corporation, presented the views of business and industry data users as a representative of the U.S. Chamber of Commerce. Ms. Naymark said the ACS “is vital to economic development and wise government and business decision-making.” She described Target’s use of census data to locate new stores and remodel existing ones, to offer merchandise and advertising to match neighborhood characteristics, and to plan for work force needs. Ms. Naymark said the ACS would provide more accurate data than the census long form because of its timeliness, and she urged Congress to fund full implementation of the ACS over the long term. Ken Hodges, Director of Demography, Claritas, told subcommittee members that, “Support and even enthusiasm for the ACS are growing in the private sector,” for whom frequent data at the neighborhood level is a “top priority.” Mr. Hodges acknowledged some concerns about the quality of ACS estimates (including coverage of the population in ‘group quarters’), but said many of the concerns also applied to the census, and that the ACS offered greater opportunity to improve the reliability of the data. Mr. Richard Ogburn, Principal Planner, South Florida Regional Planning Council, testified that the ACS would “support the more effective allocation of scarce public resources” through improved assessment of need and better targeting of resources. He noted that South Florida’s population grew by 21 percent in each of the last two decades. Without more timely data from the ACS, Mr. Ogburn said, local governments and businesses must divert programmatic resources to collect information on local characteristics. The Council analyzes area trends and conditions to support land use and natural resource planning, emergency preparedness, transportation, affordable housing, and other important services, Mr. Ogburn told the panel. House Democrats question no-adjustment decision for 2010: In a letter to the current and former top Democrats on the House census oversight subcommittee, Census Bureau Director C. Louis Kincannon reaffirmed his agency’s decision not to plan for a statistical adjustment of the 2010 census for reapportionment and redistricting purposes. Instead, Mr. Kincannon wrote, the bureau would evaluate coverage improvement programs used in Census 2000 that “had a positive influence on the differential undercount,” to see if they can further reduce the gap in accuracy between the count of Whites and of racial and ethnic minorities in 2010. Census 2000 programs that yielded “a higher percentage of enumerations of historically undercounted groups” included the Be Counted Program, Service-Based Enumeration (e.g. soup kitchens, homeless shelters), Coverage Improvement Follow-up, and the Update/Enumerate and Urban Update/Leave operations, the director said. Rep. William Clay and Rep. Carolyn Maloney (D-NY) wrote to Mr. Kincannon in May, saying they were “surprised” by his announcement at a National Academy of Sciences census panel meeting that the Census Bureau would not statistically correct the 2010 census numbers used for congressional apportionment and redistricting. The decision, the legislators wrote, is “the only aspect of the 2010 design that has been finalized.” Reps. Clay and Maloney noted that the undercount of African Americans was three percent higher than the undercount of non-Hispanic Whites in 2000 – “only a modest improvement,” they said, from the 3.9 percent ‘differential undercount’ in the 1990 census. The legislators asked about the Census Bureau’s plans to measure the accuracy of the 2010 census and to correct for overcounting. In his June 30th response, Director Kincannon said his comments at the National Academy confirms “a widely held and often articulated conclusion based on an exhaustive analysis” of the post-census survey in Census 2000. The Census Bureau learned that it could not both carry out an accuracy measurement survey and analyze the results in time to meet the legal deadline for delivering redistricting data to states, Mr. Kincannon said. The bureau also has concluded that “science is insufficiently advanced to allow making statistical adjustment … of a successful decennial census in which the percentage of error is presumed to be so small that adjustment would introduce as much or more error” than it fixes. In addition to coverage improvement programs to improve accuracy, the Census Bureau is looking at ways to clarify who should be included on the census form, the director noted. Better presentation of residence rules and an “undercount detection question” to identify households where some residents may have tenuous attachments, are under consideration. The bureau also is researching why the census double-counts people and housing units. “Improving and measuring overall coverage and differential coverage in the census are primary goals of the 2010 census,” Mr. Kincannon wrote. Upcoming news: The Census Bureau is expected to release the results of a test of voluntary versus mandatory response to the American Community Survey in the coming days. The test, conducted at the request of Congress, took place over the course of several months earlier this year. Congress asked the Bureau to assess the consequences of making response to the survey voluntary. Response to the census (short and long forms) is required by law. We will report on the results of the evaluation when they become available. Upcoming meetings: The next Decennial Census Advisory Committee meeting will be October 9 – 10, 2003. The five Race and Ethnic Advisory Committees will meet jointly October 1 - 3, 2003. The Census Bureau has not yet published agendas for the meetings. ----------------------- Census News Briefs are prepared by Terri Ann Lowenthal, an independent consultant in Washington, DC. Please direct questions about the information in this News Brief to Ms. Lowenthal at 202/484-3067 or by e-mail at terriann2k(a)aol.com. Thank you to the Communications Consortium Media Center for posting the News Briefs on the Census 2000 Initiative web site, at www.census2000.org. Please feel free to circulate this information to colleagues and other interested individuals. -- Ed Christopher Planning Specialist Resource Center Federal Highway Administration 19900 Governors Drive Olympia Fields, Illinois 60461 708-283-3534 (V) 708-574-8131 (cell)