Introduction

BACKGROUND

The 2000 census will provide a great deal of auxiliary information (i.e., information in addition to that requested on census forms) on how well various 2000 census operations were carried out. This information is relevant to the task of our committee, namely, assessing how modifications to the design of the 2000 census might work in 2010. This auxiliary information comes from several sources. First, in monitoring the 2000 census operations, the Census Bureau will be collecting information, often on a temporary basis, as part of several data systems that govern various census processes. For example, one data system, the Operational Control System 2000 (OCS 2000), will be monitoring nonresponse follow-up. (Other data systems are described below.)

Second, the Census Bureau is also obtaining useful information for planning the 2010 census from three additional sources:

  • evaluation studies—All major systems used in the 2000 census will have associated evaluation studies to determine how successfully they were carried out. These are often based on quality control systems put in place for this purpose. Others have suggested that the use of decennial census administrative records, especially cost data that result from various internal audits that are conducted, may be useful as an evaluation tool. The panel has not explored this matter in any depth to date.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 7
DESIGNING THE 2010 CENSUS: First Interim Report Introduction BACKGROUND The 2000 census will provide a great deal of auxiliary information (i.e., information in addition to that requested on census forms) on how well various 2000 census operations were carried out. This information is relevant to the task of our committee, namely, assessing how modifications to the design of the 2000 census might work in 2010. This auxiliary information comes from several sources. First, in monitoring the 2000 census operations, the Census Bureau will be collecting information, often on a temporary basis, as part of several data systems that govern various census processes. For example, one data system, the Operational Control System 2000 (OCS 2000), will be monitoring nonresponse follow-up. (Other data systems are described below.) Second, the Census Bureau is also obtaining useful information for planning the 2010 census from three additional sources: evaluation studies—All major systems used in the 2000 census will have associated evaluation studies to determine how successfully they were carried out. These are often based on quality control systems put in place for this purpose. Others have suggested that the use of decennial census administrative records, especially cost data that result from various internal audits that are conducted, may be useful as an evaluation tool. The panel has not explored this matter in any depth to date.

OCR for page 7
DESIGNING THE 2010 CENSUS: First Interim Report experiments—For each recent decennial census, the Census Bureau has used the unique environment created by the census to conduct several experiments. These experiments are used to examine the feasibility and value of a potential change in the next census. a master trace sample—The Census Bureau is planning on collecting a master trace sample that will provide comprehensive information for a sample of housing units as to how each was processed through each step of the 2000 census. THE PANEL'S CHARGE The Panel on Research on Future Census Methods was convened to assist the Census Bureau in selecting and evaluating alternative designs for the 2010 census, focusing on plans for the census tests and for research and analysis. The panel was convened in 1999 to help the Census Bureau determine whether the auxiliary data planned to be collected in the 2000 census are adequate to support the test plans and research and analysis. Because many of the plans concerning census data systems and experiments were nearly final by the panel's first meeting in June 1999, the panel focused on minor modifications for the census data systems and the census experiments. Since the plans for the master trace sample were not final, the panel was able to consider a range of possible improvements. Finally, the panel received only limited information on the evaluation studies to be used for assessing the quality of the 2000 census and so could not assess how they might be modified to provide more information for evaluating alternative designs for the 2010 census. The panel's charge for this first interim report required predicting, to the best of its ability in late 1999, the alternative designs that will be the most promising candidates for the 2010 census. (Given the testing cycle, this means predicting the primary alternatives for the 2010 census as viewed in 2004.) To make this prediction, one must forecast, among other things: the degree of responsiveness of the U.S. residents to mail questionnaires from the government, with or without an advertising campaign; the concerns of the U.S. residents about privacy and confidentiality; the number of persons expected to have affiliations with more than one household; the quality, content, and availability of administrative records; the degree of general access to the Internet; and the effect of other technological innovations. Each of these dynamic factors—and undoubtedly others—that are

OCR for page 7
DESIGNING THE 2010 CENSUS: First Interim Report extremely difficult to predict will influence the effectiveness of different designs for the census in 2010. Lott and Keller (1999) have provided a list of important considerations relevant to the ultimate design of the 2010 census. Using this and other information,1 the panel believes that alternative designs for the 2010 census can be usefully distinguished on four dimensions: the degree to which sampling will be used in the 2010 census, including a postenumeration survey, sampling for coverage improvement,2 and possibly sampling for nonresponse follow-up; the degree to which administrative records will be used, for master address list improvement, for help with specific groups, for coverage improvement, for coverage measurement, as sources of census data (including long-form information), and for imputation for item and unit nonresponse; the degree to which greater use of technology could benefit either data acquisition or data dissemination; and interaction with the American Community Survey (ACS), which includes both improvements to the master address file and eliminating the need for a census long form. Although these dimensions are useful as a starting point, the panel recognizes that they do not fully describe all the promising alternative designs for the 2010 census. THE PANEL'S ACTIVITIES TO DATE In the time available, the panel has been able to schedule two full panel meetings, for which the Census Bureau provided presentations on future census designs, the American Community Survey, census 2000 address list 1   Three presentations at the panel's first meeting were especially helpful: the future of sampling in the census, by Alan Zaslavsky; the future use of administrative records in the census, by John Czajka; and anticipated advances in technology relevant to the census, by panel members Daryl Pregibon and Michael Meyer. 2 Coverage improvement programs are census programs that attempt to address potential omissions in the census through additional efforts to include specific types of housing units or specific types of individuals, such as additional follow-up of housing units initially identified as vacant and the parolee/probationer programs in 1990.

OCR for page 7
DESIGNING THE 2010 CENSUS: First Interim Report formation, census 2000 experimentation, census 2000 evaluation plans, plans for the master trace sample, and census 2000 auxiliary data systems. In addition, a working group of the panel has met with Census Bureau staff to discuss the initial plans for the master trace sample and statistical aspects of the plans for the experiments to be carried out during the 2000 census. Finally, a working group has participated in two conference calls with Census Bureau staff concerning the auxiliary data systems for the 2000 census. In light of the relatively recent convening of our panel, we commend the Census Bureau for its conscientious efforts in assisting the panel in its work by providing excellent background material and technical presentations that have brought the panel up to speed. In addition, we commend the Bureau for getting such an early start on planning for the 2010 census, though, as mentioned above, the lack of detail on the specific evaluation studies that was provided limited the panel in its ability to review that important component of the auxiliary data collected in 2000. Realizing that the briefings of our panel drew resources away from other important census activities as April 2000 fast approaches, we are very grateful for this assistance. We look forward to future panel meetings and to the opportunity to carry out our charge as the results of the 2000 census are analyzed and planning for 2010 moves forward. STRUCTURE OF THE REPORT The panel's activities and deliberations have resulted in this report, with analysis and recommendations to the Census Bureau. The next three sections cover the key topics noted above: the master trace sample; the 2000 census research experiments, especially the experiment on administrative records; and the 2000 census data systems. In addition, although the main charge to the panel for this report focused on the plans for auxiliary data collection for the 2000 census, the broader charge is to examine alternative designs for the 2010 census, and consistent with the broader charge, the final two sections cover other 2000 census issues and planning for the 2010 census and the ACS survey.