ADMINISTRATIVE RECORDS: INTRIGUING PROSPECTS, FORMIDABLE OBSTACLES
Rapid expansion of the coverage, content, and computerization of administrative records containing information about persons, families, addresses, and housing units is a fact that is receiving increased attention from both statistical agencies and data users. How and to what extent can enhanced uses of these records supplement and perhaps even replace the more traditional methods for direct collection of population and housing data? What potential exists for significantly reducing the cost of producing small-area data and obtaining data more frequently?
It would be unnecessarily limiting to look at such uses of administrative records solely within the context of the decennial census. The decennial census is part of a population and housing data system that also includes continuing and periodic sample surveys of households and persons and a program of intercensal population estimates that already makes use of several kinds of administrative records. Administrative records maintained by program agencies at all levels of government are used by the statistical units in some of those agencies to produce valuable statistical data.
So far there have been only limited uses of administrative records in the decennial censuses (private communication, Susan Miskura, 1993). For future censuses, the panel is exploring their possible uses either as the primary source of census data (as is now done in some European countries) or as a significant adjunct to traditional census methods, to improve coverage, content, or the efficiency of census operations and to evaluate census results. Both this panel and the Panel on Census Requirements have chosen to study possible uses of administrative records within a broader context. Thus, we are considering possible uses of administrative records in the 2000 census and the testing of such uses in 1995, and we are also looking at censuses beyond 2000 and at other components of the U.S. demographic data system.
We begin this chapter by summarizing prior recommendations about uses of administrative records from our panel's December 1992 letter report (Committee on National Statistics, 1992) and from the interim report of the Panel on Census Requirements (Committee on National Statistics, 1993). We then discuss the underlying policy and broad technical issues that we believe are critical to the success or failure of efforts to make more effective use of administrative records for demographic data. We review and comment on the current status of plans for using administrative records in the 2000 census and for testing such uses in 1995. We also look beyond the 2000 census toward a long-term program of research and
development that could lead to a census based primarily on administrative records in 2010 or to other significant uses of administrative records for small-area population and housing data. Finally, we identify some questions about administrative records that we expect to consider and address in our final report.
PRIOR FINDINGS AND RECOMMENDATIONS
In the panel's letter report of December 14, 1992, to the director of the Census Bureau (Committee on National Statistics, 1992), we agreed with the Census Bureau's conclusion that a census based primarily on administrative records was not a feasible option for 2000 (Bureau of the Census, 1992a). A primary reason given by the Census Bureau for this conclusion was that the race and ethnicity information needed from the census to comply with requirements of the Voting Rights Act is not currently available for most persons from any single administrative records source or combination of sources. In the view of the panel, however, "The primary issues surrounding the two administrative record designs . . . are not necessarily voting rights data, but privacy, coverage, geography, and the whole range of content" (Committee on National Statistics, 1992:6)
Our panel believes that the Census Bureau should be doing more to explore emerging possibilities for expanded uses of administrative records in its demographic data programs. Specifically, the panel recommended that the Census Bureau:
initiate a separate program of research on uses of administrative records, not directly tied to the 2000 census, focusing primarily on the 2010 census and on current estimates programs.
undertake a planning study to develop one or more detailed design options for a 2010 administrative records census (a prospectus for such a study was included as an attachment to the panel's report);
seek the cooperation of federal agencies that maintain key administrative record systems in undertaking a series of experimental administrative records mini-censuses, including one concurrent with the 2000 census, and related projects; and
give priority to some use of administrative records in the 2000 census for those purposes for which such use is still feasible, such as coverage and content improvement and coverage evaluation.
The Panel on Census Requirements, in one of three formal recommendations in its interim report, specifically endorsed the above recommendations. In addition, the text of the requirements panel's interim report makes clear that that panel supports increased exploitation of administrative records for intercensal population estimates, starting immediately, both as an end in itself and as a means of gaining experience that can lead to their expanded use in the decennial census. In discussing its future activities, the requirements panel expressed the belief ". . . that it is important to examine the potential of administrative records to complement or perhaps substitute
for all or part of the decennial census" (Committee on National Statistics, 1993:30). The requirements panel plans to examine operational, legal, administrative, and privacy issues related to such uses of administrative records.
Finally, in Appendix B of its interim report, the requirements panel included, without endorsement or recommendations but with an invitation for comments, a short paper, "Developing New Sources for Small-Area Data." The system described in that paper would exploit current information technology and would make extensive use of administrative records. It might eventually replace the decennial census as the main source of small-area data, providing data much more frequently than once every 10 years.
The system would use administrative records from all levels of government in three ways. Initially, records from a particular system could be used on a stand-alone basis to summarize and map data for geographic areas already identified in the system. For example, a system containing addresses with ZIP codes in its records could be used to map data to the ZIP-code level. As a second step, these same records could be linked to a geographic database to produce data at the individual block or block face level. A third step would link records from different sources at the individual address level in order to undertake analyses that would not be possible with aggregate data from each source. There would, of course, be important questions of confidentiality protection associated with each type of use, especially for linkages at the individual address level.
As an illustration of what might be done, the paper describes Statistics Canada's program that provides annual data on tax filers and sources of income, based on tax records, for relatively small areas. A recent comparison of the data from this source with the 1991 Canadian census showed that the tax records data covered 97 percent of the census population and 99 percent of the population in families (Standish et al., 1993). This particular illustration belongs in the first category described above: it does not require any linkages to a geographic database or to records from other administrative sources. Comparable U.S. studies are discussed following recommendation 4.6 (below).
BASIC REQUIREMENTS FOR MORE EFFECTIVE USE OF ADMINISTRATIVE RECORDS
The panel believes that many statistical uses of administrative records are appropriate and have important advantages. Such records can play an important part in meeting needs for basic demographic data, especially for small geographic areas. In this section we discuss major policy and technical issues that must be addressed in an effective long-range effort to develop effective uses of administrative records.
Effective use of administrative records by the Census Bureau requires a legal right to access, the establishment of close and mutually beneficial ongoing relationships between the Census Bureau and the custodians of administrative records, and reasonable assurance of continued access. By a legal right to access, we mean at least the absence of legal prohibitions on access for statistical uses, and, preferably, positive statutory recognition of statistical uses as a permissible secondary use of the records. The Census Bureau probably has greater legal access to administrative records than any other U.S. statistical agency, but its access is by no means universal, especially with regard to systems maintained at the state level, for which access is controlled by state laws.
Legal access is necessary but not sufficient for effective use. Access in specific instances is often arranged only with great difficulty. These difficulties will continue unless both statistical agencies and the custodians of record systems develop new ways of thinking about statistical uses of administrative records. The custodians need to regard satisfying the statistical requirements of the nation as one of the responsibilities on which they will be judged, not as an inconvenience or an intrusion on their territory. There has to be some flexibility on the part of both statistical and program agencies in adapting their operating procedures and schedules to meet basic statistical needs. There must also be willingness by the statistical agencies to assist administrative agencies in every possible way, for example, by providing technical support to the latter for the production of small-area tabulations of administrative records, to the extent that this can be done while ensuring a one-way flow of administrative records for statistical uses.
Statistics Canada's access to administrative records, which makes possible the data system described in Appendix B of the Requirements Panel's Interim Report, may provide a useful model. For the United States, a 1976 policy statement of the Social Security Administration (U.S. Department of Health and Human Services, 1976:17) is relevant:
The operation of the social security system produces a vast and unique body of statistical information about employment, payrolls, life-time earnings histories, retirement, disability, mortality and benefit claims and payments.
The Social Security Administration has an obligation to develop these data according to the best scientific standards and with maximum economy and minimum delay, and to publish them in a form useful both to the program administrator and to social scientists generally. It also has an obligation to encourage the linkage of these data with other bodies of statistical information, and to make the data available for research uses by other organizations, subject always to careful safeguarding of the confidentiality of information relating to individuals.
The panel notes the Census Bureau's initiative in organizing a July 15, 1993, Interagency Conference on Statistical Uses of Administrative Records and its plan to hold a similar meeting involving custodians of state record systems. We urge the participants to take advantage of this beginning by moving to establish a regular mechanism for closer collaboration between custodians and users of administrative records.
Effective use of Internal Revenue Service (IRS) records in the decennial census and other Census Bureau programs requires a close working relationship between the two agencies. In the past, the two agencies have sometimes had difficulty in agreeing on their respective roles and in developing and maintaining smooth working arrangements. Given the importance of these programs to the entire federal statistical system, these difficulties suggest a need for involvement by the Statistical Policy Office of the Office of Management and Budget (OMB).
The Census Bureau can only commit itself to substantially greater reliance on administrative records if it can have reasonable assurance of continued access. As an illustration of the type of problem that can occur, Standish et al. (1993) note that, as a result of changes in Canada's system of family allowances, some of the information about children that had been used in the statistical program (described above) will no longer be available. It is also important that proposed legislation requiring the establishment of major new administrative records systems be carefully monitored to ensure that possibilities for important statistical uses of the records are recognized and are not unnecessarily foreclosed. New systems established in connection with health care reform might offer a less controversial alternative to other major federal record systems as a primary source of population data.
Recommendation 4.1: The Statistical Policy Office in the Office of Management and Budget should recognize statistical uses of administrative records as one of its major areas of responsibility and should assume an active role in facilitating more effective working relationships between statistical and program agencies and in tracking relevant legislation.
No plan for expanded statistical uses of administrative records can succeed unless it has the acceptance of the people who provide the data to the program agencies. Greater use of administrative records could reduce the number of separate occasions on which individuals are asked to provide information about themselves to the government, and it also has the potential for substantially reducing the cost of censuses. The key issues, however, are consent and confidentiality. Do people accept the use of information about themselves for statistical purposes that are not directly related to the purposes for which they supplied the information? Should they be able, as individuals, to prevent such uses? Will the confidentiality of their data be adequately protected? Effective use of administrative records for census evaluation,
coverage improvement, and supplementation of content may require that identifiers—such as full name, date of birth or Social Security number—be collected in the census and entered into census electronic files, along with addresses. Will this be acceptable? These are not easy questions. We note that many of them have been explored in the report of the Panel on Confidentiality and Data Access (Duncan et al., 1993).
The panel believes that it is necessary to proceed with public debate about the ethical, legal, and policy issues associated with statistical uses of administrative records. The Census Bureau has had numerous contacts with privacy advocates in connection with previous censuses and has informed the panel that it plans future discussions that will focus specifically on statistical uses of administrative records. However, there has been a tendency on the part of the agencies involved to back away from a public debate for fear that calling attention to these questions might lead to discontinuance of important existing activities, such as the use of tax return and Social Security data in the Census Bureau's intercensal population estimates programs. We believe it is time to face these questions directly. Such a debate is likely to be more productive if it focuses on specific uses of administrative records, such as the implications of an administrative records census in 2010, rather than on broad philosophical questions.
Surveys and focus group interviews can help provide background for public discussions of these issues. Some information on taxpayers' opinions about possible uses of their data in the census of population and for other statistical purposes was collected in a survey conducted for the IRS, the 1990 Taxpayer Opinion Survey. Neither the report nor the planned public-use microdata files from that survey have yet been released. When they are, the panel expects to analyze the data for information about the public's views on these issues.
Recommendation 4.2: The Census Bureau should initiate a systematic process of consultation and research to explore the attitudes of the public, political representatives, and other opinion leaders about the use of administrative records as an integral part of the census. Previous consultations and existing research, such as the yet-to-be-released 1990 Taxpayer Opinion Survey, should be taken into account.
Several of the central themes that are discussed in Chapter 1—including address list development, record linkage research, the need for long-term planning and the need for greater interagency cooperation and coordination—are critical to making better use of administrative records in the decennial census and for other programs to produce small-area population and housing data. The first two of these are technical requirements, and in this subsection we discuss briefly their particular relevance to the use of administrative records.
As stated in Chapter 1, the panel believes that a geographic database that is fully integrated with a master address file is a basic requirement, whether for a traditional census or for one making greater use of administrative records. To be cost-effective and available for noncensus uses, the system should be continuously updated. Its coverage should be extended to rural areas, and conversion of rural-type addresses to street addresses should be promoted by all means possible. Appendix B to the interim report of the Panel on Census Requirements explains clearly how such a system could serve as the keystone for using administrative records data to provide small-area data more frequently and inexpensively. We recommend (Recommendation 1.1) that the transition to a continuous, integrated system begin now, at least for the areas in which the 1995 census test is to be conducted, so that it can receive its first major tryout in the 1995 test.
Record linkage—that is, the identification of records belonging to the same unit, either within a single data set or in two different data sets—is critical to enhanced uses of administrative records, whether in the context of a full administrative records census or a more traditional design. If different administrative data sets are to be used to improve the coverage of the master address file or of persons at known addresses, duplicates must be identified and eliminated. Uses of administrative records to supply missing data or to evaluate the coverage and content of the census enumeration all require some type of record linkage for persons or addresses.
Record linkage, like any element of census data collection and processing, is subject to error. The Census Bureau and other organizations have developed effective techniques for linking large data sets, but many aspects need further research and development as the techniques are applied in specific circumstances. How can the inputs be standardized to facilitate linkages? Over how wide an area should initial computerized matches be undertaken? What are the best keys (e.g., name, address, date of birth, or Social Security number), alone or in combination, and what are the costs and other considerations for capturing these items in the computerized files for a census or evaluation survey? Additional research on these questions is a prerequisite to success in making more effective use of administrative records. Testing should be carried out in conjunction with the 2000 census and the tests leading up to it and also in separate initiatives to explore the possibility of a 2010 census based primarily on administrative records.
Another technical requirement for enhanced uses of administrative records is knowledge about the quality of the data that they can provide. How well do administrative record systems, individually or in combination, cover the target population for a census? Recent work in the United States (Sailer et al., 1993) and in Canada (Standish et al., 1992) indicates that well over 90 percent of the respective countries' enumerated populations can be identified in their tax systems, not supplemented by any other source. How much could this coverage be improved by adding records from other systems, and how would the coverage of subgroups defined by geography or other characteristics (differential coverage) be affected, both in absolute and relative terms?
Besides coverage, the accuracy and relevance of data available from
administrative records need to be considered. For example, suppose there are administrative data for persons whose census responses are incomplete. Is enough known about the quality of these data to make an informed choice among alternatives: further nonresponse follow-up, imputation based on similar persons, or substitution of the administrative data? To what extent could tax data be used either to evaluate or substitute for income data collected in a census? What are the implications of conceptual differences? Opportunities and resources should be sought to pursue questions like these.
USE OF ADMINISTRATIVE RECORDS IN THE 2000 CENSUS
During the remainder of this decade, the primary goals for expanded uses of administrative records should be: to experiment in the 1995 census test with the uses considered most promising for the 2000 census and, based on the results, to make appropriate uses of administrative records in the 2000 census; and to begin an active continuing program of research and development for a possible 2010 census based primarily on administrative records and for other uses of records in demographic data programs. These two streams of activity should proceed in parallel, and they must be carefully coordinated. Uses of administrative records in the 2000 census are discussed in this section. The following section presents our views on research and development that focuses on the 2010 census and other uses.
In our earlier letter report, we accepted the Census Bureau's judgment that an administrative records census would not be possible for 2000, but we recommended that the Census Bureau give priority to some uses of administrative records in the 2000 census for purposes for which such use is demonstrated to be feasible. Possible uses in the census can be classified into four broad groups:
Measures to improve coverage. Coverage improvement measures divide into those aimed at improving the master address file and those aimed at improving the coverage of individuals. The former can be used throughout the decade, or at least prior to the census, to ensure a good starting address list. The latter need to use records current at the time of the census and will tend to focus on administrative sources that are rich in data on difficult-to-enumerate subpopulations. Both kinds of uses of administrative records could be made across the board or for a sample of blocks as part of a built-in adjustment process designed to produce a one-number census. Administrative records would be a logical element of a CensusPlus or SuperCensus design (see Chapter 2) in which close to 100 percent coverage is sought for a sample of areas, or they could serve as one of the sources in a dual-or triple-system estimation design.
Measures to improve content. Administrative records have already been used to some extent for evaluation purposes—for example, in record checks with tax data to assess income reporting. A next step could be to use them as a source of data to replace data that are missing due to nonresponse or data that failed edit tests. The
final level would be to use administrative records as the initial source of data for some variables, with some form of follow-up for missing data cases as necessary.
Measures to improve operational efficiency. The use of administrative records as a source of telephone numbers is one example of use for operational efficiency. Administrative records data could be used prior to the census to identify hard-to-enumerate areas for which special enumeration methods might be appropriate.
Measures to evaluate the census. Uses of administrative records for evaluation will depend very much on how the 2000 census methodology develops with regard to integrated coverage processes that are designed as part of a one-number census. To the extent that administrative records are not used to improve content, they could be used to evaluate content.
It is now clear that the 1995 census test will not be a test of one or more of the 14 designs proposed at the beginning of the selection process; rather, it will be a test of features taken from several of the 14 designs. Many of the features selected for testing will represent significant changes from the 1990 design. Although informal discussions have identified some of the features that are receiving strong consideration, the Census Bureau has not yet released a detailed proposal identifying and describing the uses of administrative records that it proposes to test.
Rather than recommend specific features for testing, the panel wishes at this time to make three broad recommendations concerning uses of administrative records in the 1995 census test. We assume that the 1995 test sites will include both urban and rural areas and that they will be determined in fall 1993, with enumeration in the spring of 1995.
Recommendation 4.3: As part of the 1995 census test, the Census Bureau should construct an administrative records database for the test sites.
The administrative records database for the 1995 census test should be designed to serve two objectives: (1) testing uses of administrative records that appear to be promising for the 2000 census, and (2) obtaining information, especially about the population coverage of major administrative record systems, that will be useful in exploring the feasibility of a census based primarily on administrative records in 2010 (for more discussion of this goal, see the following section). In order to serve both purposes, the database for the 1995 test sites should include address and person information extracted from both federal and local record systems. Federal systems, particularly those of the IRS and Social Security Administration, are known to cover a large proportion of the total population.
If the Census Bureau fails to take advantage of the 1995 test to start exploring the possibilities of a census based largely on administrative records in 2010, there will be a danger of repeating the pattern of the current and earlier decennial census cycles, in which consideration of the administrative records option was begun too late to make it a serious contender for the subsequent census.
Recommendation 4.4: The Census Bureau should establish the testing of record linkage procedures as an important goal of the 1995 census test.
For this purpose, it will be necessary to collect and capture electronically one or more of four potential matching keys: full name, date of birth, address, and Social Security number. Address and either name or birth date should be captured for all persons enumerated, and all four keys should be captured for a sample of persons (with the Social Security number being treated as a voluntary item), so that their relative effectiveness as matching keys can be evaluated. The provision of Social Security number should be clearly identified as voluntary on questionnaires and in follow-up interviews.
Recommendation 4.5: In preparation for uses of administrative records in the 1995 census test, detailed negotiations between the Census Bureau and the other relevant agencies should begin immediately, with the involvement of the Statistical Policy Office of the Office of Management and Budget (see also Recommendation 4.1).
In addition to the content and delivery dates of records to be made accessible to the Census Bureau, the negotiations could cover possible modifications to standard administrative record formats for the test areas during the relevant periods and tabulations that the other agencies might make of their own records for comparative evaluation purposes.
USE OF ADMINISTRATIVE RECORDS IN 2010 AND BEYOND
A successful long-range effort to define and develop new uses of administrative records for demographic data requires a strategic plan, embracing a full array of censuses, current surveys, population estimates, and small-area data from program records as a system. Such a plan cannot succeed if it is developed unilaterally by the Census Bureau. Within the federal statistical community, planning must involve other statistical agencies and OMB's Statistical Policy Office. A broad range of data users must be consulted, and the custodians of major administrative record systems must be actively involved.
As detailed above, the panel recommended in its letter report that the Census Bureau initiate a separate program of research on uses of administrative records, focusing primarily on the 2010 census and on current estimates programs (Committee on National Statistics, 1992). It also recommended a planning study to develop one or more detailed design options for a 2010 administrative records census and initiation of a series of administrative records mini-censuses.
The Census Bureau has indicated its intentions to start these activities in fiscal 1995 if resources permit. The panel believes that an earlier start would be desirable so that the needs of the long-range program can be taken into account in planning for
the 1995 census test. The 1995 census test and the 2000 census will provide valuable opportunities to gain experience that will facilitate uses of administrative records in a 2010 census and in other programs. Once prototype designs have been developed for a 2010 administrative records census—an objective of the ''HOW Study'' that we recommended (Committee on National Statistics, 1992)—the following type of development program can be envisioned:
pilot test(s) to construct a census purely from administrative records in a defined and limited geographic area—primarily to learn of the problems;
test(s) as in (1), but in conjunction with a conventional census test to allow evaluation of coverage;
test(s) that incorporate follow-up activities to investigate missing addresses and households for which the administrative data fail to meet standards for completeness and consistency—primarily procedural tests;
test(s) as in (3), but in conjunction with a conventional census test to allow evaluation of coverage (which will require careful design to ensure that the administrative records test and the conventional census do not affect each other);
full stand-alone dress rehearsal(s) of an administrative records census in chosen areas, using the procedures found to be most effective in steps (1) to (4).
The 1995 census test would present the first opportunity for a type (2) test. Tests of types (3), (4) and (5) could be undertaken in conjunction with the final tests for the 2000 census and the census itself. The main point we wish to make here is that, in planning for the 2000 census, the Census Bureau should make a serious effort to incorporate testing of possible administrative records uses for the 2010 census and other purposes.
The 1950 decennial census provides a model for this kind of long-range planning and experimentation. Self-enumeration, which was extended to most of the country in 1960, was tested in experimental areas in the 1950 census. The 1950 census also saw the first use of an electronic computer (UNIVAC I) for data processing on a limited basis; by 1960 most of the processing was done on computers (Goldfield, 1992). In planning for future uses of administrative records, it is necessary to break out of the 10-year planning cycle that always looks only to the next census. If this does not happen, intriguing prospects will not become real gains.
The Census Bureau staff that have been assigned to explore administrative records uses have displayed imagination and competence. Much has been learned about the characteristics of administrative record systems in both the public and private sectors, and this information is being systematically documented in the Census Bureau's Administrative Record Information System. These accomplishments hold promise for the future, provided the two-stream approach (described above) for the 2000 census is adopted and sufficient resources are committed soon to a serious ongoing development program.
Recommendation 4.6: The Census Bureau should establish a formal program of long-range research and development activities relating to expanded use of administrative records for demographic data.
As recommended by the Panel on Census Requirements (Committee on National Statistics, 1993), uses of administrative records in the current population estimates program should be a part of this effort.
Recent research by the Statistics of Income Division of the Internal Revenue Service has produced an estimate of 1990 U.S. population by combining data from income tax returns and information documents. The IRS estimate was equal to 98.7 percent of the unadjusted 1990 census count and 97.1 percent of the adjusted count. (The 95 percent confidence interval for the latter percent is from 96.2 percent to 98.0 percent.) State estimates exceeded 92 percent of the adjusted census counts for all but six states (Sailer et al., 1993).
Although one should not make too much of these promising but preliminary results, they do suggest the possibility that data from IRS and Social Security Administration (SSA) record systems can potentially play a much greater role than they do now in providing population counts or estimates, with some information on demographic and economic variables. The data from the IRS/SSA systems could be an important component of improved current population estimates, of a continuous measurement system, as envisaged in Design Alternative Recommendation (DAR) #14 (Bureau of the Census, 1993b) and the prototype described by Alexander (1993), and of new small-area data systems as described in Appendix B of the requirements panel's report (Committee on National Statistics, 1993).
Ideally, these possibilities should be explored jointly, with close collaboration at all stages, by the Census Bureau, the IRS and the Social Security Administration. Until now, however, the IRS's research on the development of population counts has been undertaken by the Statistics of Income Division with little input from the Census Bureau. The Statistics of Income Division has limited resources for this kind of research. Under the assumption that the Census Bureau will succeed in obtaining the resources for long-range research and development referred to in Recommendation 4.6 (above), we suggest that the Census Bureau consider the possibility of using some of these resources to fund contract work by IRS on statistical uses of IRS/SSA records. The objectives of such work would be specified in the context of the long-term needs of federal demographic data programs, with emphasis on further exploration of potential statistical uses of annual files that integrate data from tax returns and information documents. The methods and approaches to be pursued at this time would be left largely to the discretion and ingenuity of IRS. Full documentation and wide availability of the results would be important requirements for this work. We hope that the Statistical Policy Office in OMB could play an active role in facilitating this kind of interagency cooperation.
In our judgment, the successful development of IRS and SSA records for demographic data purposes depends on those organizations' having a recognized and explicit role in a cooperative process. The above approach would represent a first
step in this direction. We think there is a substantial likelihood that the results of such contract work would be so promising as to make the value of these data sources for federal demographic data systems self-evident. If that does occur, we believe that any legislative, organizational, or changes needed for such uses can be addressed.
The panel is mindful of the difficulties that confront efforts to use administrative records more effectively. Public acceptance is critical. A wider understanding of the difference between statistical and nonstatistical uses of records is essential, and people must be convinced that the confidentiality of records used for statistical purposes will be fully protected. We hope to have available soon the database from the 1990 Taxpayer Opinion Survey so that we can improve our understanding of these issues. Content and user requirements for small-area data are also primary considerations. The lack of consistent and complete race and ethnicity information in administrative records systems is a special problem and one to which we plan to give further attention. We urge the Panel on Census Requirements to seek a clearer picture of what kinds of data are essential for very small areas and therefore cannot be provided by large sample surveys.
The nation's demographic data programs cannot commit to reliance on administrative records as a primary or major source of data unless continued and timely access to the key administrative data sets is assured. Access requires both clear legal authority and broad agreement throughout the executive and legislative branches of the government on the desirability of using the records for statistical purposes. As indicated in the chapter title, these are formidable obstacles. But we believe the potential benefits—in terms of data that are of equal or better quality, cost less, and are available more frequently—are great. The possibilities should be thoroughly explored, and methods sought to overcome the obstacles.