« PreviousContinue »
SECTION 1-THE NATURE OF THE FRAMEWORK
Role of the Framework
programs which they administer and evaluate. In many cases, the needs for information cut across the interests of several agencies. Consequently, a number of interagency committees, ad hoc task forces, and informal consultations are involved in the development of many of the statistical programs. Nevertheless, it is frequently difficult for individuals representing their respective agency needs or interests to identify and influence key statistical programs of potential use. This document, especially with its orientation toward agency missions and functional statistical areas, is designed to facilitate improved coordination among agencies and individuals with interest in specific data series.
A Framework for Planning U.S. Federal Statistics for the 1980's is designed to provide an overview of the U. S. Federal Statistical System through discussion of subject areas, statistical agencies, and important related crosscutting issues. It will be a framework for more detailed discussions concerning the improvement of existing statistical programs, efforts to meet important gaps in available information, and recommendations for discontinuance of low-priority statistical efforts in each of the areas. The document is not a final policy statement on any particular issue, but is presented to provide background and guidance for resolving specific issues.
In its final published form in 1979, the Framework will reflect substantial effort by Federal statistical agencies, various advisory and user groups, selected policymakers, and the Office of Federal Statistical Policy and Standards (OFSPS). All statistical agencies have been encouraged to prepare detailed individual program plans which move beyond the routine management documentation required for budget making and program design. The bibliography identifies a number of planning documents which are available from individual agencies.
The final document will be the product of a multiyear effort which was formally instituted in January 1975, under the auspices of the Interagency Committee on Statistical Programs and Policy (now the Federal Council on Statistical Programs and Plans). This group is comprised of heads of major U.S. Federal statistical agencies. It created a planning task force which set into motion a program for agency review of problems, programs, and related issues.'
This document is intended to serve several distinct and important objectives. They are as follows:
This Framework focuses on multipurpose and large-scale statistical programs. It does not attempt to be exhaustive in reviewing or relating the full range of detailed statistical activities of the various agencies. In its present form, this Framework may appear to be cumbersome and overly detailed; for other readers, however, it may fail to encompass a significant portion of Federal statistical activities. Its focus on multipurpose statistical programs and largescale statistical activities provides a framework for examining priorities among the more expensive statistical activities of the Federal Government. Although no attempt is made to specify a rank
'For background statements on the planning process, see Joseph W. Duncan, “Developing Better Long-Range Plans for Federal Statistics,” Statistical Reporter No. 75-4, October, 1974; Robert W. Raynsford, “The Interagency Statistical Planning Effort, 1975," Statistical Reporter, No. 76-3, September, 1975; Paul O'Neill, "OMB's Role in Planning and Coordination of Federal Statistics," Statistical Reporter, No. 76-11, May, 1976.
See also Joseph W. Duncan, “Priority Setting in the Coming Decade (Survey Linkage and Integration,” Statistical Reporter, No. 77-7, April 1977; Joseph W. Duncan, “Developing A Framework for Planning U.S. Federal Statistics, 1978-1989,” The American Statistician, August 1977, Vol. 31, No. 3; Joseph W. Duncan (ed.), Future Role of Government Statistical Services, A Review of the Operation, Development, and Future Role of Statistical Services in Major Countries of the West and Eastern Europe, Report of the Seminar of the Chief Statisticians of Member Nations of the Economic Commission for Europe (Oxford: Pergamon Press, 1978).
The U.S. Federal Statistical System is highly decentralized. Individual statistical efforts are frequently initiated by agencies to relate to the basic
U.S. Federal statistical agencies. It does attempt to create a framework within which individual surveys can be better evaluated by stating the missions and areas of responsibility of individual agencies. Further, it tends to focus on problem areas, with the result that some of the excellent ongoing programs receive relatively limited attention.
ordering of various statistical projects, the Framework is intended to be helpful in identifying new initiatives which should be undertaken and the consequent improvements to the statistical system which would result. In the final chapter, a staged program is presented to demonstrate the types of things which should be done in various time periods throughout the 1980's. Further, the recommended improvements are frequently designed to permit discontinuance of lower quality or more narrowly conceived statistical programs. It is anticipated that a substantial amount of further public review and comment on the Framework will provide a wide range of suggestions concerning specific priorities and programs which may be implemented in the future.
Character of the Plan
The Framework is designed to highlight several important crosscutting issues. These are presented in Section IV. These issue papers are subject to substantial improvement as more detailed research is undertaken and as various policy groups examine these specific issues. Those topics which have been selected for consideration in the crosscutting statistical issues section have been identified on the basis of discussions with the individual agencies and with important user groups. They are not intended to represent the full range of crosscutting issues; they simply highlight some important activities which are especially crucial in the current evolution of the U.S. Federal Statistical System.
The Framework is designed to be an extremely flexible document. In effect, its format is that of a looseleaf collection of concept and issue papers. An attempt has been made to ensure that the presentations are developed at a consistent level of detail and organization. Early drafts of these materials were circulated for comment and criticism in order to ensure that the first published versions are adequate to represent the diversity of issues and programs considered. It is anticipated that individual sections will be revised extensively throughout the 1978-1980 time period. It will continue to be subject to revision as priorities change, as individual problems are solved, and as difficulties arise in the implementation of the proposals which are presented. Current plans call for a full revision to be published at the end of 1979, and it is expected that individual chapters will be updated in subsequent years.
This Framework is not intended to serve as a budget document. It is published under the authority of Section 103 of the Budget and Accounting Procedures Act of 1950 which gives the Office of Federal Statistical Policy and Standards responsibility for planning and coordinating the U.S. Federal Statistical System. Obviously, the proposals to institute new programs, to discontinue old programs and to undertake significant research and evaluation are all activites which have budgetary implications. In basic design, the Framework is presented as background to those specific decisions and not as an indication that those specific decisions have been made or necessarily will be made within a specified time frame.
The Framework is explicitly designed to review individual statistical agencies, individual statistical programs, and crosscutting issues with the purpose of clarifying important interactions in the decentralized U.S. Federal Statistical System. Section IIOrganization and Operations of U.S. Federal Statistical Agencies—contains important recommendations concerning principles which will lead to improvements in the present organization. For most programs, however, organizational change is seen as an evolutionary phenomenon, and the solution of the specific problems presented in this Framework is
It is anticipated that the presentation of the materials in this Framework, especially in the first integrated presentation, will be helpful to the general public for evaluating the current status of Federal statistics and for making specific recommendations for improvements. It is clear, as noted above, that most statistical agencies will direct much of their attention to the detailed programs for which they have responsibility. Continuing feedback from the general public to the Office of Federal Statistical Policy and Standards is therefore encouraged and helpful to the overall process. The Framework is expected to generate specific suggestions from the public and user groups for several years to come.
Since the Framework attempts to encompass the major multipurpose and large-scale statistical programs of the U.S. Federal Statistical System, it does not, in any way, seek to detail the wide range of individual series or specific surveys undertaken by
expected to be feasible without significant reorganization. In some areas, reorganization and centralization of certain activities will enhance the feasibility of achieving suggested improvements. Specific reorganization proposals are expected to come from the President's Reorganization Project. They will be incorporated in the 1979 revision.
This Framework is presented in the context of the 1980's. Whereas a number of ideas in the Framework are provocative, they are not presented as a “futuristic" view of the U.S. Federal Statistical System. Rather, this planning Framework represents a practical, evolutionary approach to building upon the ongoing programs of the 1970's and the unresolved issues associated with those programs. A perspective on the 50 years of evolution in methodology and program development which occurred between 1926 and 1976 is presented in a companion volume entitled Revolution in United States Government Statistics, 1926-1976.
statistical programs. In recognition of the critical role of all three of these groups, therefore, the development of the Framework continues to involve, in an active and dynamic way, all of these groups.
Though all three groups have been involved in the process of preparing and reviewing the Framework, it is recognized that there will be many issues and problems which are not addressed in the Framework. The major general-purpose statistical programs, especially those with interagency implications, are given greatest priority since the most significant contribution of the Framework itself is to enhance the coordination process among these agencies.
Thus, the Framework concerns itself with areas of overlapping interest among Federal statistical agencies. It is recognized that there also exist many areas of concern which fall entirely within one agency's sphere of interest, but these areas have not been included in the Framework, in order to provide reasonable limits to its scope.
It is expected that, as a result of the participatory process utilized in developing this Framework, there will be a number of related documents prepared offering alternative views, more detail concerning specific proposals, and consideration of issues which are not treated fully in this document. The Office of Federal Statistical Policy and Standards will continue to serve as a focal point for identifying such documents, but users of this Framework are encouraged to seek out these alternative views on the specific issues discussed in this report.
Planning Is A Participatory Process The development of the Framework for Federal statistics is a highly participatory process involving many important users and producers of statistics. The planning process itself was designed to include soliciting agency views, building on existing agency planning programs, and utilizing readily available resources in the Office of Federal Statistical Policy and Standards (which has overall responsibility for planning and coordinating the Federal Statistical System). The ultimate resolution of statistical issues involves an important collaboration among the providers of the information (the respondents to various Federal inquiries), the collectors and analyzers of that information (the statistical and analytical agencies), and the private and public decisionmakers who incorporate that information into their decision processes. All three sectors — the providers, the statistical agencies, and the decisionmakers-are essential to the development of the Federal Statistical System. Obviously, no statistics would be available if respondents failed to cooperate in surveys. The statistical agencies, in their central role of creating effective collection mechanisms, providing quality control, and compiling and analyzing results, have important creative contributions to make as a result of this direct involvement in the statistical process. Finally, the decisionmakers who utilize the information are able to provide important insights into the level of quality and timeliness required to assure appropriate policy use of statistical information. Further, in many cases, the decisionmakers directly influence the allocation of
Keys to Integrated Statistical Systems While Section V provides a summary of the major sections of the Framework, it may be helpful to identify several keys to an integrated statistical system at the outset.
Brief comments on each of these areas are presented as an introduction to the detailed discussion which is contained later in this report.
Standard Concepts and Classifications
In the highly decentralized statistical system of the United States, it is evident that there are many data collection efforts which use differing definitions, time periods, and sample populations for various important statistical series. While variation is inevitable, especially in narrow-purpose statistical programs, the basic recommendations in this Framework are designed to extend the idea of standard concepts and classifications into a broad set of general-purpose, large-scale statistical efforts which have interagency
implications and uses. The development of standard concepts and classifications is a complex process involving compromise and considerable technical ingenuity. Recommendations for achieving this are presented in appropriate sections throughout this report.
implementing improvements in statistical programs. Recommendations for extending and improving the existing network of technical interagency working groups are made throughout the report in recognition of the vitality of this mechanism in the creative evolution of the Federal Statistical System.
General-Purpose Collection Efforts
In an attempt to enhance the implementation of standard concepts and classifications and to extend the usefulness of individual collection efforts, considerable attention is focused throughout this report on the development of general-purpose data collection efforts which meet multiple needs. A basic premise of this Framework is that extended generalpurpose collection efforts can increase the quality of U.S. statistical information in a cost-effective manner. This cost-effectiveness is directly related to replacing costly, narrow-purpose statistical efforts with general-purpose collection efforts whose total cost may often be less than the combined cost of the discontinued programs and which may simultaneously yield improvements in the quality of the data provided.
With the emphasis in this Framework on the needs for and uses of statistical information, it is evident that the Federal Statistical System benefits greatly from public advice received from those individuals who utilize Federal statistics in their analytical and decisionmaking processes and who provide the essential data to the Federal Government by completing the various reporting requirements. In some agencies the advisory mechnism is well established and effectively utilized. In other agencies important user groups should be involved more directly and effectively than is presently the case. Consequently, several recommendations for improving the relationship between public suppliers and users of information and statistical collection and analytical agencies are presented.
Central Policy Agency
The Framework is viewed as a starting point for considering individual statistical programs and developing responses to the priority problems identified. It is therefore essential to have a set of government decisionmakers who are able to accurately reflect the policy concerns of the various agencies and who can articulate the priorities for statistical data in the individual subject areas. A number of policy committees with this purpose are identified in the discussions presented in Section III. Recommendations for creating and improving several policy committees are also set forth in appropriate sections.
For over 40 years there has been a centralized office with responsibilities for improving the planning and coordination of Federal statistical activities. The present document has been developed by the organization with that continuing responsibility, the Office of Federal Statistical Policy and Standards in the U.S. Department of Commerce. It is evident as a consequence of developing this planning document that the interagency perspective provided by such an agency is an essential component in ensuring the appropriate development of statistical programs of high utility. Several recommendations are presented for assignments which should be undertaken by interagency groups under the leadership of this Office. Even though the control function of the central statistical planning agency is limited in its directive aspects, the leadership which can be provided through careful working relationships with the agencies is especially important to the continued development and implementation of improvements in the U.S. Federal Statistical System.
Technical Interagency Working Groups
The key to a strong Federal Statistical System is the people involved in the day-to-day design, collection, processing, and analysis of Federal statistics. The interagency working groups that bring these talented individuals together are essential for