THE COMPILATION OF MANUFACTURING STATISTICS Frank A. Hanna DOCUMENTS DEPARTMENT SEP 16 1960 LIBRARY UNIVERSITY OF CALIFORNIA U.S. DEPARTMENT OF COMMERCE BUREAU OF THE CENSUS Robert W. Burgess, Director Washington : 1959 SOCIAL SCIENCES For sale by the Superintendent of Documents, U.S. Government Printing Office, нало 16H23 SOCIAL SCIENCES FOREWORD As industry's products, processes, and institutions have become more complex, the business community, Government agencies, and the public have sought an ever-increasing volume of statistics on manufacturing. At the same time demands have been placed upon manufacturers to file a sharply growing number of statistical and administrative reports. In this setting much more attention is now paid than formerly to the value of information relative to the cost of providing it. The breadth of scope, completeness, accuracy, amount of detail, and frequency of data must be weighed against the cost to the Government and the respondent in compiling them. As a consequence, new techniques have been introduced for determining what information is needed, what is available in industrial records, and how to collect and process it. Documentation of both the conceptual and the procedural aspects of these issues has been spotty. Previous authors have described various phases of the subject but their accounts have been limited in scope or remote in time. It seemed to us at the Census Bureau, therefore, that an up-to-date and comprehensive account of the major steps in compiling industrial statistics was needed. Accordingly, this book was prepared at the request of the Census Bureau to augment the information available on the programs and activities leading up to and including the publication of industrial statistics. In it Dr. Hanna has set forth the content, interrelations, and compilation techniques of a broad system of statistics. He has provided a reference that gives some insight into the origin, purpose, and processing problems of particular and their relation to other industrial surveys and the system as a surveys, whole. As a comprehensive and current description and discussion of the compilation of industrial statistics, the book can be regarded as a partial fulfillment of an important responsibility of the Bureau. In a single volume it supplements and integrates the relative meager notes, definitions, and appendixes ordinarily included with published statistics. There are other uses it is expected to satisfy these relate to the instruction of people who plan to compile industrial statistics rather than to use them. Such instruction has always been a major concern in the training of the Bureau's new employees. Similarly, the treatment in a single document of all principal phases of the compilation of industrial statistics should facilitate the Bureau's technical assistance to representatives of less ex perienced nations that are preparing to expand their statistical activities. Finally, the availability of a definitive book telling the what, why, and how of industrial statistics should prove helpful to college students, respondent manufacturers, research organizations, trade associations, and others concerned with this growing field. The early discussions of the need for a book such as this began shortly after World War II, when the Bureau of the Census was attempting to unify its current and benchmark industrial statistics. Dr. Hanna, then advisor on program planning in the Bureau's Industry Division, played an important role in this statistical coordination. Again, in 1951, Dr. Hanna served as Project Manager for Technical Assistance Mission No. 77 sponsored by the Organization for European Economic Cooperation. The purpose of this mission, the members of which included many leading European statisticians, was to study industrial censuses in the United States. The review and evaluation of countless documents in the files of the Bureau of the Census and the forging of selected pieces of information into a balanced and cohesive description has been a major task, the completion of which depended importantly upon Dr. Hanna's familiarity with the development of industrial statistics in the United States and his acquaintance with the organization and operation of the Bureau of the Census. It is indeed fortunate that he was willing to undertake the assignment. MAXWELL R. CONKLIN, Chief, Industry Division, Bureau of the Census. PREFACE Many different approaches to a subject as broad as manufacturing statistics are possible. I have tried to give a description of the entire process of compiling statistics on manufacturing from conception to publication and distribution, paying particular attention to the problems which in the middle and late fifties were of some consequence: these are the materials needed to appraise the probable effects of most sources of error on manufacturing statistics. The choice of this approach has guided the selection of discussion topics, and has delimited and set the levels of detail for the treatment of each topic. The concepts which are controversial and the processes which are likely to affect the character of the statistics have been accorded greater attention than those which it seems safe to take for granted. About equal space is devoted to conceptual problems and to the problems inherent in the various processing operations. Most of those who have occasion to use manufacturing statistics have a general knowledge of the content of the Census of Manufactures and of many of the processes by which statistics are collected and compiled. Substantive and processing problems can be discussed fruitfully, however, only in detailed and precise terms. This requirement has necessitated a series of guesses as to how much of statistics and statistical processing is common knowledge and how much must be either described or left to the reader's imagination, and as a guide I have tried to keep in mind the needs of a graduate student or inexperienced professional worker. Those with greater familiarity with statistical processes will, I suspect, welcome having a detailed description of these processes published, so that the relevant parts can be referred to but need not clutter their own arguments. This book was written during 1957-58, when the 1954 Census was substantially completed. The planning for the 1958 Census, although it was still preliminary and subject to change, is mentioned here in a number of instances, since even a tentative decision for 1958 often helps to point up a difficulty encountered in the 1954 Census or in earlier ones, or is useful for indicating the current thinking about a continuing problem. Some references are made to the Census of 1939 and earlier ones, but almost all of the discussion is focused on the postwar period. Although more attention is devoted to the 1954 Census than to others, there are many references to the 1947 Census, since it incorporates most of the changes made since 1939. Alaska, which was included in the Census of Manufactures as early as 1899, became a State after this volume was written. |