and industry, under either alternative assumption of enactment or nonenactment of the pending legislation. In other words, the estimates are based upon a total of five such limited annual censuses under the existing schedule and a total of eight under the schedule proposed in the pending bill. The estimated costs of such a limited census in either case has been placed at $300,000 per annum, divided equally between the census of manufactures and the census of business. We think it is fair to assume this even division of cost in view of the probability that it will be necessary to enumerate completely the manufactures of the United States, though with a limited schedule, while sampling methods might be employed in the case of the larger number of establishments which are involved in a census of business. While we do not suggest the substitution of other specific figures for any of those appearing in the Census Bureau's statement, it is our opinion that its estimates are in general too high. This opinion pertains particularly to the census of manufactures, for which an average cost in excess of $2,000,000 for a single census is assumed if taken quinquennially, or slightly less than two million if taken biennially. We believe that in both cases the estimates should be closer to one and one-half million dollars. In this connection we should like to present a comparison of the estimates in the statement under discussion with figures which were furnished by the Bureau of the Census to the Bureau of the Budget in 1937. These earlier figures have been so adjusted as to provide comparable estimates for major items of cost during a decade for censuses of business, mines, and manufactures, on the assumption that the pending bill is enacted: Estimated costs of censuses of business, mines, and manufactures during a decade From this table it appears that the present estimates by the Bureau of the Census of its requirements under the terms of the pending legislation are four and one-half million dollars lower than the Bureau would have submitted 4 years ago. However, its present estimate of the cost as a single item of the proposed quinquennial census of manufactures is more than double that of 4 years ago. There are certain other striking differences in the two sets of figures. However, all estimates of this kind are of necessity highly speculative and lacking in uniformity when they are made without the benefit of precise specifications. The cost of any census necessarily varies with its scope and this depends in turn upon the extent and variety of the demands for information which it undertakes to satisfy and upon the magnitude of available funds. Your next request was that we submit a list of Federal agencies engaged in gathering statistics. The list which I attach has been compiled from data now being assembled for the sixth edition of the Directory of Federal Statistical Agencies, and has been corrected to November 10, 1941. Some of the agencies included do not necessarily themselves collect statistics but make use in their statistical work of data collected by others. While I believe this list to be as complete and as accurate as it is possible to compile, I am sure that it will bear out my comment during the hearing that it is very difficult to define a "statistical agency." The larger part of the statistical work of the Federal Government is carried on by administrative agencies and a list of statistical agencies must necessarily include the names of administrative agencies whose primary duties are not statistical. Your final request was that we give the committee information concerning savings in the statistical expenditures of other Federal agencies resulting from actions taken by the Division of Statistical Standards. As I implied in my testimony, such an estimate is both difficult to make and likely to be misleading. While it is a relatively simple matter to count the number of questionnaires which have been abandoned by Federal agencies in accordance with our recommendations the major service that we aim to render consists in promoting coordination and improvement of the statistical work of the executive agencies. Inefficient and ill-coordinated statistical work by these agencies, resulting in a lack of needed information, can be far more costly in the long run both to Government and to the public than an occasional superfluous questionnaire. Moreover, when the staff of the Division has intimate and constant contacts with other Federal agencies it is able to learn of proposed statistical investigations before they reach the stage of formal proposals. It is in a position to discourage unwise investigations before they are formulated and to participate in the development of desirable inquiries while they are still in their formative periods. The problem is not unlike that of placing an evaluation upon the services of a supervising architect who reviews building plans prepared by other persons. He may reject some plans entirely, substitute alternatives for others and in some cases build suitable structures from very poor beginnings. Now and then he can point to definite, tangible savings resulting from substitutions of materials or designs, or from his calculation that the buildings as planned would have been unnecessarily large. The worth of his services could not, however, be adequately represented by the magnitude of such savings. Moreover, to the extent that the developments of architectural plans had been under his direct guidance from their inception, his opportunities for such savings would disappear; the more effectively his supervisory functions were performed from the beginning, the fewer gross errors and misjudgments would remain for him to correct. Similarly, it would not be inaccurate to say that the Division of Statistical Standards is successful in its work to the extent that it early removes the possibility of showing tangible savings. An estimate of direct savings resulting from the withdrawal of questionnaires after our review of them was presented on April 25, 1939, in hearings before the Committee on Expenditures in the Executive Departments, House of Representatives, Seventy-sixth Congress, first session, on H. R. 5917, "A bill to amend the act establishing the Central Statistical Committee and the Central Statistical Board," as follows (p. 36): "In addition to the definite economies which the general operations of the Board have brought about, it believes that the review of plans for statistical inquiries has also resulted in very substantial savings. Plans for inquiries are frequently submitted to the Board before costs have been estimated. It is impossible, therefore, to estimate precisely the savings which have resulted from the abandonment of projects following the Board's review. However, as set forth in section II, nearly 300 report forms, out of a total of 3,400 formally reviewed by the Board, were withdrawn by the submitting agencies for extensive revision or abandonment. No wholly defensible, detailed estimate of the resulting savings can be made, but it is the concensus of the Board's review staff that an expense of at least half a million dollars to the Federal Government has thus been avoided. This estimate does not include savings effected by the shortening of questionnaires and by other simplifications of procedure resulting from Board recommendations." The experience cited in this excerpt was that of the Central Statistical Board during its lifetime prior to the date of the hearing. In response to your request we have attempted to arrive at a corresponding estimate during a recent 4 months' period of operation by the Division of Statistical Standards. Limiting the estimate to questionnaires completely withdrawn following the Division's review, and taking into account such considerations as proposed frequency of use, probable number of respondents to be covered, probable costs of tabulation, printing, analysis, etc., we arrived at the opinion that the direct savings to the Federal Treasury would represent at least a million and a half dollars. An estimate such as this takes no account of the far more numerous instances in which questionnaires have been shortened or procedures improved. In discussing the usefulness of this Division I should like again to emphasize the need for constructive improvements in the Nation's store of information. I believe that our legitimate claim to the rendition of public service rests upon such constructive improvements far more than upon the direct dollar savings which also stand to our credit. I hope that in preparing this statement I have correctly understood your desires. I shall be glad to provide so far as I can any further additional information which you would like to have. Sincerely yours, STUART A. RICE, Assistant Director, in Charge of Statistical Standards. Estimated costs relating to those parts of census work affected by the proposed bills S. 1627 or H. R. 5232 for the 10-year period 1942-52 Total for limited annual benchmark statistics collected for the 8 noncensus years for Manufactures and Business.. $17,849, 600 $16, 811, 200 1 1,500,000 Grand total.... 20, 249, 600 18, 311, 200 1 The proposal has been made and much consideration given to collect restricted annual statistics in intercensal years for Manufactures and Business. This item is inserted to cover such a possible addition LIST OF FEDERAL STATISTICAL AGENCIES 1 Compiled by Division of Statistical Standards, Bureau of the Budget, (corrected to November 10, 1941) Executive Office of the President: Bureau of the Budget. National Resources Planning Board. Office of Government Reports. Office for Emergency Management: Division of Defense Housing Coordination. Office of Civilian Defense. Office of the Coordinator of Inter-American Affairs. Office of Defense Health and Welfare Services. Office of Facts and Figures. Office of Lend-Lease Administration. Office of Price Administration: Consumer Division. Price Division. Division of Accounting Analysis and Review. Defense Economics Department. Transportation Department. Office of Production Management: Bureau of Research and Statistics. Division of Civilian Supply. Division of Contract Distribution. Labor Division. Materials Division. Priorities Division. Production Division. Purchases Division. Transportation Division. 1 This list includes some agencies which do not themselves collect statistics but which make use of the data collected by other agencies. Economic Defense Board: Office of Export Control. Selective Service System. Office of the Secretary. Agricultural Adjustment Administration. Agricultural Marketing Service. Bureau of Agricultural Economics. Bureau of Agricultural Chemistry and Engineering. Bureau of Animal Industry. Bureau of Dairy Industry. Bureau of Entomology and Plant Quarantine. Bureau of Home Economics. Commodity Credit Corporation. Commodity Exchange Administration. Farm Credit Administration. Farm Security Administration. Federal Crop Insurance Corporation. Office for Agricultural Defense Relations. Rural Electrification Administration. Soil Conservation Service. Surplus Marketing Administration. Department of Commerce: Office of the Secretary. Bureau of the Census. Bureau of Foreign and Domestic Commerce. Bureau of Marine Inspection and Navigation. Weather Bureau. Department of the Interior: Office of the Secretary. Bituminous Coal Division. Bureau of Mines. Bureau of Reclamation. Consumers' Council Division. Fish and Wildlife Service. General Land Office. Geological Survey. Grazing Service. National Park Service. Petroleum Conservation Division. Office of Indian Affairs. Office of the Petroleum Coordinator for National Defense. Department of Justice: Office of the Administrative Assistant to the Attorney General. Antitrust Division. Bureau of Prisons. Federal Bureau of Investigation. Immigration and Naturalization Service. Department of Labor: Bureau of Labor Statistics. Children's Bureau. Division of Public Contracts. Wage and Hour Division. Women's Bureau. Department of the Navy: Naval Contract Distribution Division, Office of the Under Secretary of the Department of the Treasury: Bureau of Customs. Bureau of Internal Revenue. Comptroller of the Currency. Division of Monetary Research. Division of Research and Statistics. Division of Tax Research. Foreign Funds Control. Procurement Division. Department of War: Contract Distribution Division, Office of the Under Secretary of War. Army and Navy Munitions Board. Office of Chief of Engineers. Matériel Division of the Army Air Corps-Wright Field, Dayton, Ohio. Federal Loan Agency: Office of the Administrator. Federal Home Loan Bank Board (including Home Owners' Loan Corporation.) Federal Housing Administration. Reconstruction Finance Corporation. Federal Security Agency: Office of the Administrator. Civilian Conservation Corps. National Youth Administration. Office of Education. Social Security Board: Bureau of Employment Security. Bureau of Old-Age and Survivors Insurance. Bureau of Public Assistance. Bureau of Research and Statistics. United States Public Health Service. Federal Works Agency: Office of the Administrator. Public Roads Administration. Public Works Administration. United States Housing Authority. Work Projects Administration. Board of Governors of the Federal Reserve System. Federal Communications Commission. Federal Deposit Insurance Corporation. Federal Power Commission. Federal Trade Commission. Interstate Commerce Commission. Maritime Labor Board. National Labor Relations Board. Railroad Retirement Board. Securities and Exchange Commission. Tennessee Vellev Authority. United States Civil Service Commission. United States Maritime Commission. Veterans' Administration. The CHAIRMAN. A good many of us on this committee have felt rather favorably toward having the Bureau of the Census act as the sole agency with authority to issue questionnaires and collect this information, to eliminate all the other agencies completely from going into this field. Of course, as you have indicated, in the Bureau of Internal Revenue there would be opposition to that. There is opposition to the proposal coming to me by mail from other agencies of Government with which you are familiar, because they have indicated that they have contacted you and that approval has been given to write to the chairman of the committee. We feel that if this authority |