reports involved in management, operation, and evaluation of Federal programs constitute the bulk of the reports. Collection of data for the longestablished economic and social statistics impose minimal reporting burdens on respondents, although minor exceptions offer opportunities for some economies. In some of the newer statistical programs where Federal activity has expanded rapidly, serious paperwork problems are evident. But the overriding problem in reducing the burden of Federal statistics lies in the mechanisms for planning and coordinating the Federal statistical system."

In summarizing its analysis of paperwork required for statistics, the Commission stated:

In the full picture of all Federal reports, however, statistical reports probably play an even smaller role than depicted by the OMB and GAO figures because of the importance of nonstatistical forms issued by exempt agencies (chiefly the Internal Revenue Service and the financial and banking regulatory agencies)....

By any test...statistics play a comparatively minor role in overall Federal reporting. Statistical surveys accounted for about 26 percent of the forms, but only 13 percent of the estimated hours spent in reporting and less than 10 percent of the annual responses in the inventories for June 30, 1976."

While statistical reports are only a small part of total governmental reporting, the very size of the statistical establishment and its decentralization in many government agencies indicates that it is essential that continual scrutiny be given to problems of reporting burden, especially for voluntary statistical surveys where the respondent is not receiving a direct benefit as a result of program participation. While it is true that all citizens benefit from improved statistics which make it possible to have enlightened policy, it is often difficult, in a specific case, for the respondent to appreciate the value of his contribution. Consequently, if overall governmental demands for information are viewed as excessive, the voluntary statistical inquiries are likely to be among the first to be ignored, especially in comparison with those reporting burdens associated with applying for a Federal grant or benefit or with required reporting on the use and application of such a benefit. Therefore, it is essential for the statistical system to make every effort to minimize reporting burden and to utilize, to the ful

10 Ibid., pp. 40-41. "Ibid., pp. 40-41.

lest extent possible, those data which are available from other administrative sources.

In the next section of this review, a number of the specific recommendations of the Commission on Federal Paperwork relating to statistics will be presented in relation to the Framework itself. The interested reader is encouraged to review the full report of the Commission entitled Statistics since it provides considerable elaboration of the rationale behind the recommendations as well as additional background information about the problems under discussion. Further, many other Commission reports in topical areas have recommendations relating to statistical programs. Since the Commission's report is widely available, the following section will be somewhat abbreviated. The recommendations are discussed below, in the order in which they are presented in the Commission's report on Statistics.

Discussion of Commission on Federal Paperwork Recommendations for Reducing Reporting Burden in Statistical Agencies

In this section the actual wording of the Commission on Federal Paperwork recommendations will be presented in italics. The discussion which follows is a perspective in the context of the overall Framework except where quotes indicate background or elaborating material from the Commission's report itself.

Recommendation No. 1. The Director of the Office of the Management and Budget should revise the staffing and other resources applied to statistical coordination in a strong central coordinating unit for the Nation's statistical system.

This recommendation became somewhat out of date just as the Commission report was being issued. This is acknowledged in the "Foreword" to the Commission report on Statistics in which it is noted that Presidential Reorganization Plan No. 1 of 1977 for reorganization of the Executive Office of the President which transferred statistical policy functions from the Office of Management and Budget to the Department of Commerce would "affect some of the recommendations which the Commission made in this report. Accordingly, the Commission made some observations on issues which it believed should be considered as the reorganization plans developed."2

12Ibid., p. 1.

After noting that the drive to reduce the burden of Federal paperwork should not slacken during the reorganization effort, the "Foreword" stated:

That a study of the U.S. "statistical system" should be undertaken immediately, with the full cooperation of the Federal agencies and professional groups, looking to potential ways to improve the functions of planning, coordination, and operation of the system. Attention should be paid, first, to adequate staffing; to more effective use of the focal agency principle with strong departmental staffs; to the regular, structured participation of the Government agencies and of users of statistics in the planning process; to provision for regular continuing professional review of the quality and of the methodology used in major U.S. statistics by such an agency as the Committee on National Statistics; to ways to build a more effective professional career system and training program for U.S. Government statisticians, as in the United Kingdom; and to the possibility of a greater degree of concentration in the collection of statistics, especially within related areas.

The Commission has recommended that the staffing pattern of the present Statistical Policy Division be reviewed. It is clearly inadequate to perform its assigned functions effectively. When the new plan becomes effective, the Commission would recommend to the Secretary of Commerce that this Statistical Policy staff be greatly strengthened."3

The sense of these recommendations is consistent with actions underway in early 1978. First, a major review of the Federal Statistical System is being undertaken as part of the President's Reorganization Project under the auspices of the Statistical Improvement Project. Materials being developed as part of A Framework for Planning U.S. Federal Statistics for the 1980's serve as important background to that project which is considering a number of the specific points mentioned in the "Foreword" to the Commission's report on Statistics.

Second, the Department of Commerce in its 1979 budget request asked for 10 additional positions to augment the 15 positions forming the core staff of the new Office of Federal Statistical Policy and Standards. This is a first step in strengthening the capabilities of the Office to deal with the planning responsibilities which are the focus of the Commission's first recommendation. It is further anticipated that following review of this preliminary integrated draft of the Framework and consideration of the President's Reorganization Project recommendations, that a number of more specific steps will

be taken to provide for stronger coordination and planning of the statistical system.

Recommendation No. 2. The Office of Management and Budget should assign coordinating responsibilities to focal agencies for specific subject-matter areas, such as health, education, agriculture, and energy.

This recommendation is closely aligned with Recommendation No. 1 since it has the objective of further strengthening the coordinating capacity within the Federal Statistical System. The reorganization of the clearance process within the Office of Management and Budget which followed Reorganization Plan No. 1 included an initial set of proposals to decentralize some of the clearance responsibilities to focal agencies, thereby implementing at least in part Recommendation No. 2. The Commission's report on The Reports Clearance Process discusses and describes the clearance activities in selected agencies in relation to this focal agency concept." In addition, the Office of Federal Statistical Policy and Standards, through established interagency committees, has been working to strengthen the focal agency role in each of the named areas. The individual chapters of the Framework dealing with these subject areas provide further details concerning developments in those areas. The Commission noted in its report on Statistics that "clearance is not only a control mechanism, but it is part of OMB's planning and coordinating effort."""

It should be noted that Reorganization Plan No. 1., though formally separating the Federal Reports Act responsibilities and the Section 103 responsibilities of the Budget and Accounting Procedures Act concerning planning and coordinating statistics, was followed by subsequent actions in which the statistical clearance function was redelegated to the Office of Federal Statistical Policy and Standards. Thus, the inherent control function associated with clearance was retained in an integrated fashion insofar as Federal statistical programs are concerned. Further examination of this separation is expected during the period following implementation of these new procedures. At the time this preliminary report was prepared, it was clear that there is general agreement that the planning and coordination function must be strengthened. Steps are being taken to assure that such developments will occur. As noted in the Commission report on Statistics (pages 34-35), the Framework program is an important tool in such planning and coordination and that continues to be the case following the reorganization.

"The Reports Clearance Process, pp. 37-40; 51-54 "Statistics, p. 39

Recommendation No. 3. Statistical service units should be established to provide technical assistance to the Congress and to Federal agencies in the planning and design of statistical studies and in the development of guidelines for awarding contracts and grants for statistical studies.

This recommendation is closely aligned with Recommendation No. 1 concerning the staff resources available in the central coordinating unit. As the Commission notes: "SPD (the predecessor of the Office of Federal Statistical Policy and Standards) has never been adequately staffed to provide technical advice and assistance on statistical problems to those Federal agencies which need such help.""

While the staff level has not been adequate to undertake a proactive program of technical assistance, the staff members of the Office of Federal Statistical Policy and Standards do provide technical assistance in the design of statistical programs when specific problems are brought to their attention by the agencies. The establishment of the Statistical Policy Coordination Committee, a cabinet-level committee, is designed to provide a tool for gaining more awareness of the needs for such technical assistance. Further, the Federal Committee on Statistical Methodology is reviewing standards for contracting and for providing technical assistance in statistical programs which are not undertaken by the major statistical agencies. Hence, it is anticipated that in the near future some direct improvement will result in relation to the third recommendation of the Commission.

It should be noted at this point that all three of these recommendations are directed to strengthening the planning and coordination function so that duplication will be minimized, and reporting burden will therefore be reduced. Even technical assistance relates to that objective since the technical assistance frequently is of the form of designing more efficient inquiries or more efficient exploitation of existing data. Both of these activities have the impact of reducing reporting burden while at the same time improving the comparability of Federal statistics.

Recommendation No. 4. The Bureau of the Census should recommend to State governments that they tablish central registries of the status and boundaries of incorporated places, now incorporations, and related changes.

The development of current information concerning city boundaries and related information

Ibid., p. 39, parentheses added to text.

has been a perennial problem related to State and local statistics. A first step in improving the State and local data information base was the development of improved procedures for intercensal population estimates. The volume of boundary changes in incorporated places is substantial. It is clearly a local responsibility to maintain updated information on such changes.

In early 1978, an investigative task force of the House Appropriations Committee was exploring a series of recommendations for developing more consistent mapping procedures and for establishing a digitized mapping system which would be available at all agencies. Further, the U.S. Department of Agriculture was exploring the feasibility of a potential system for improved land use data. Both of these activities will contribute valuable information toward the solution of the problem being addressed by this Commission recommendation. Its exact impact on reporting burden, however, is unclear since only a limited amount of statistical reporting burden relates to queries concerning detailed local-area information.

Recommendation No. 5. The Bureau of the Census should make the results of the Census available more promptly; provide computer tapes that, while preserving individual response confidentiality, can be used readily by research and other groups; improve the capability for rapid special tabulations of Census data; and provide more effective assistance to groups and individuals who need assistance in the use of the data tapes.

Obviously, the more rapid dissemination of census data will have the effect of reducing the need for special surveys and reports immediately following the census period. This will contribute to some reduction in reporting burden. Current plans for the 1980 census include extensive efforts to make the data available more rapidly and to provide through State centers more sophisticated access to available information.

As noted in the chapter on accessing Federal data files, the extent to which Federal statistics can become more readily available and more easily accessible will have a substantial impact on reducing the need for special inquiries, thereby reducing reporting burden.

Recommendation No. 6. The Office of Management and Budget should examine existing intercensal reports and develop guidelines to require that the census data for 1980 and 1985 be fully utilized and that special intercensal surveys be permitted only in cases of demonstrated need for additional information.

This recommendation of the Commission on Federal Paperwork is strongly endorsed throughout the Framework. One of the most critical overall recommendations contained in the Framework concerns the use of the mid-decade census as an integrating device for social statistics. If properly planned and designed, the mid-decade census can offset the needs for many special-purpose surveys which presently occur on an intercensal basis. Additional details concerning this aspect of the Framework are provided in a final section which sets forth the overall strategy of the Framework. However, as noted in that section, the mid-decade census costs would be offset in large part by the elimination of intercensal special-purpose surveys. This cost reduction is paralleled by a reduction in reporting burden in noncensus years.

The Commission on Federal Paperwork expressed some concern about the psychological harassment of governmental inquiries. It is felt that the more publicized, more carefully explained census activities have a minimum negative psychological impact. Hence, the more integrated approach not only reduces the number of special inquiries, but hopefully reduces the psychological burden of Government questionnaires. Thus, this recommendation of the Commission will receive serious attention throughout the coming planning of the 1980 and 1985 censuses. Recommendation No. 7. The Bureau of the Census, in regard to the Census of Agriculture, should: (1) Substitute the use of sampling for a universal count wherever possible; (2) Utilize a much simpler, shorter form; and (3) Field test proposed Census forms on working farmers.

Present plans for the Census of Agriculture incorporate all of these recommendations. The forms will be shorter for the 1978 Census of Agriculture and a number of data items will be estimated from samples. Appropriate field testing has always been a part of the Census program.

Recommendation No. 8. The Secretary of Agriculture should appoint a temporary study group to study the current and anticipated needs for data on agriculture and on socioeconomic conditions in rural America, and plan a broad-gauge integrated statistical system to meet those needs.

The response to this recommendation is thoroughly discussed in the Framework chapter on agricultural statistics. Further, the Statistical Improvement Project is currently addressing some of these issues. Hence, this Commission recommendation will be the subject of further study between now and 1980, and it is anticipated that a more integrated system of agricultural statistics will

have the net effect of reducing reporting burden on the rural and agricultural sector of the society.

Recommendation No. 9. The Office of Management and Budget should establish, in conjunction with the agencies, a schedule for periodic reviews of all major general-purpose statistics of the Federal Government, by committees including outside experts, to assure that the best statistical techniques are being used and the reporting burden is kept at a minimum.

This recommendation for periodic review is parallel to the statistical audit recommendation of the Commission on Federal Statistics which reported in 1971. There has been widespread support for this general concept and, in fact, a number of individual reviews have been undertaken recently or are currently underway. For example, the Gross National Product Data Improvement Project, which completed its work in 1977; the Wholesale Price Index review and evaluation, which was completed in 1977 by the National Bureau of Economic Research; and the ongoing President's Commission on Employment and Unemployment Statistics represent outstanding examples of such reviews. It is true, however, that there is not a routine schedule of such reviews, and that it would be desirable to institute such a program. As noted in the concluding section of the Framework, a proposed review schedule is set forth, and it is hoped that general agreement on such a review schedule will be achieved.

It should be emphasized, of course, that the focus of such reviews should be improving the quality of statistics while keeping reporting burden to a minimum. Hopefully, those general-purpose statistical series which are out of date would be eliminated, with a consequent reduction in reporting burden. Frequently, however, as noted in the three studies cited above, the quality improvements may have the net effect of increasing reporting burden. In those cases, the social judgment to be made is whether the value of the improved information offsets the additional reporting burden. These are judgments which will have to be made on a case-bycase basis. Tradeoffs between improved statistics and reporting burden must be carefully evaluated before new surveys are undertaken.

Recommendation No. 10. The Bureau of Labor Statistics should develop, working with State Employment Security Agencies, a sample rotation plan whereby small employers would report for no more than 2 years at a time.

The Bureau of Labor Statistics has done some preliminary investigation of this recommendation and expects to do more feasibility testing. A stronger

program of respondent follow-up is also an important aspect of making this recommendation operational.

This completes the section on the formal recommendations of the Commission on Federal Paperwork relating to statistics. However, it should be noted that several important points, with respect to statistics, were made in other programmatic reports from the Commission. The crosscutting implications of the burdensome Federal reporting requirements in the fields of education, energy, and health care received special focus by the Commission, and in all cases, it was discovered that there were similar problem areas which relate to statistical activities. In view of the importance of these general principles (which given further support to Recommendation Nos. 1-3), the Commission's report on Statistics (pages 103-105) is quoted intact on these major common problems and proposed solutions.

Interagency Coordination

Coordination among the agencies of plans for data collection was found to be inadequate in each of these fields, so that inquiries sent to the public sometimes overlapped or contained duplicate or near-duplicate questions. No comprehensive or well-developed lists of reports or detailed data were available either in the agencies or at OMB as a guide to preventing duplicate collections at the planning and clearance stages, although work was underway on this task. Efforts to centralize or consolidate data collection, with one agency effectively acting as a "broker" for agencies with related programs, had not progressed very far.

Recommendations to improve interagency coordination hinge on strengthening an existing data or statistical center to permit development of a complete inventory of the existing data base for storage in a computer. This basic step should be coupled with the development either by the center or advisory groups of standardized terminology and definitions for the entire field. Once these two steps are completed, the Commission contemplates that the results would be used as an integral part of the clearance process, with clearance withheld for forms that contain duplicative questions or do not use the standardized nomenclature.

The next step in the improvement process deals with the coordination of the actual collection of data and would entail combining inquiries for different agencies in one survey or having one agency serve as a collection agent for the other agencies in the field.

The ultimate step toward perfecting interagency coordination is the designation of a focal agency in each field. Under this concept, the various program and statistical agencies in the field would channel material concerning their reporting requirements to the focal agency, which would then be responsible for developing an annual plan for information gathering, serve as OMB's deputy in the clearance process, and make appropriate arrangements for consolidating or coordinating collection activities. Special arrangements with respect to certain parts of this procedure would probably have to be made for regulatory agencies that have mandatory authority to obtain certain information from respondents.

In addition to the proposals for coordinating statistical operations in each of these major fields, the Commission urges that the Congress exercise restraint in legislating specific reporting requirements and consider the paperwork implications of legislation prior to enactment.

Standardization of Terminology

The lack of standardized terminology, definitions, and units of measurement prevented some agencies from using data already collected by another agency without extensive adjustments. They generally regarded adjusting the data as far less satisfactory than collecting data tailored to their own specific needs. This lack of standardization added to the burden on respondents because new and slightly different compilations were required. Data were not shared with other agencies in some cases, often grounded in the provisions of laws and regulations concerning the confidentiality of the data collected. Work has begun on developing standard terms in all of these major areas, but has been completed for a relatively small proportion of the statistics with which the agencies were concerned.

The recommendations for improving interagency coordination also would go far toward lowering the barriers to the sharing of information that arise from the lack of standardization. In the first place, available information would be more readily accessible and standardized definitions would make it more readily usable by other agencies. The focal agency concept could also permit the development of a service agency which could collect various types of information needed in the field, tabulate it, and send as much of it as necessary or desirable to each agency concerned. Before this approach could be fully implemented, however, knotty questions hinging on differences among agencies in the laws and regulations concerning confidentiality of the data collected would have to be resolved. Again, special