CHAPTER XX CENSUS OF MANUFACTURES--CONCENTRATION RATIOS Abstract.....................................XX-3 Introduction.................................XX-5 File Specifications..........................XX-9 Database Structures...................XX-11 User Notes..................................XX-13 CENSUS OF MANUFACTURES--CONCENTRATION RATIOS Abstract CONTENTS Citation XX-3 Type of File XX-3 Universe Description XX-3 Subject-Matter Description XX-3 Geographic Coverage XX-3 Technical Description XX-3 Reference Materials XX-3 Related Printed Reports XX-4 Related Machine-Readable Data Files XX-4 Availability XX-4 CITATION Census of Manufactures on CD-ROM, 1987: Subject Series: Concentration Ratios [machine-readable data file] / prepared by the Bureau of the Census. -Washington: The Bureau [producer and distributor], 1993. TYPE OF FILE Summary statistics UNIVERSE DESCRIPTION The universe of this file is companies with one or more paid employees primarily engaged in manufacturing. Manufacturing is defined as the mechanical or chemical transformation of substances or materials into new products. SUBJECT-MATTER DESCRIPTION These data are exactly the same as shown in Tables 3 and 4 of the printed Concentration Ratios report. No file corresponding to Tables 1 and 2 is provided since they are quite small. This file presents for each industry the total number of companies, the total value of shipments, and the percent of value of shipments accounted for the 4, 8, 20, and 50 largest companies. Also shown is a Herfindahl Index, discussed in the next section. Data are presented for 1987 and every prior census year back to 1935 for which comparable data are available. GEOGRAPHIC COVERAGE This file includes data only at the U.S. level. TECHNICAL DESCRIPTION File Structure: dBASE III+ MC87C4 Concentration Ratios 1. File sequence: 2-digit major industry groups for 1987, followed by 4- digit industries, with years 1987 to 1935 in descending order within SIC. 2. Available index: MC87C4S: First record for each SIC 3. Linkable label files: MC87SIC (SIC titles) REFERENCE MATERIALS 1987 Economic Censuses CD-ROM 1E, Technical Documentation. This documentation has general information, glossaries, record layouts for all files, and other reference material. One copy accompanies each CD-ROM order. Additional copies are available for $10 from Customer Services, Bureau of the Census, Washington, DC 20233. Guide to the 1987 Economic Censuses and Related Statistics. Describes the scope, coverage, classification systems, data items, and data products for each of the economic censuses and related surveys. Data comparability and uses are also discussed. Single copies are free from Customer Services, Bureau of the Census, Washington, DC 20233. Standard Industrial Classification Manual, 1987. Prepared by the Executive Office of the President, Office of Management and Budget, and available from the Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. Price is $24. RELATED PRINTED REPORTS 1987 Census of Manufactures, Subject Series: Concentration Ratios. Includes all of the data contained in this file, plus two brief tables summarizing concentration for all manufacturing industries combined. RELATED MACHINE-READABLE DATA FILES Census of Manufactures, 1987: Industry Series. Provides statistics on employment, payroll, value added, shipments, and new capital expenditures for all industries, primarily at the national level. Some historical statistics are shown for years between 1972 and 1982 for industries that changed in definition in 1987, and for which no historical data are presented in this Concentration Ratios file. AVAILABILITY This file is included on CD-ROM 1E (price $150). CENSUS OF MANUFACTURES--CONCENTRATION RATIOS Introduction CONTENTS General Description XX-5 Concentration Ratios and the Establishment Basis of Reporting XX-5 Concentration Ratios and the SIC System XX-5 Explanation For Tables XX-6 Herfindahl-Hirschmann Index XX-6 GENERAL DESCRIPTION This report provides ratios on the share of- industrial activity accounted for by the largest companies, both for the United States as a whole and for each of the 459 manufacturing industries. The first presentation of concentration ratios in manufacturing by the Bureau of the Census was made in 1939 and was based on the 1935 Census of Manufacturers. Ratios for all manufacturing industries were presented in the National Resources Committee publication, The Structure of the American Economy. In the 1937 census, concentration ratios were presented for a large sample of individual products. Beginning in 1947, with the first full census following World Ward II, concentration, concentration ratios have been presented as part of every census of manufactures. In 1954, 1958, and 1963, the Census Bureau conducted the project in cooperation with the Senate Subcommittee on Antitrust and Monopoly. The Census Bureau has continued to present this widely-used report in the censuses of 1967, 1972, 1977, 1982, and 1987 for the benefit of data users in the industrial educational, and government communities. As in previous reports, no attempt is made to interpret the significance of the concentration ratios. The ratios are not used as the basis for an analysis of either the causes or effects of concentration. The report consists merely of a presentation of the ratios themselves. Through years of usage , the term "concentration ratio" has come to mean the share of the total activity or resources of a given segment of the economy of the economy accounted for by its largest companies. In order to avoid disclosure of the operations of any individual firm, the concentration ratios based on census data are presented in groups of four companies or more, that is, the four largest, eight largest, and so forth. CONCENTRATION RATIOS AND THE ESTABLISHMENT BASIS OF REPORTING The concentration ratios based on census of manufactures data have consistently been prepared from establishment reports. Under the establishment system, the shipments of a given company are distributed among all of the industries in which it has establishments. In tables 1 and 2 of this report, which reflect the activity of the largest companies in the industrial sector as a whole, the establishment approach means that only the manufacturing establishments of these companies are taken into account in determining their size. In tables 3 and 4, a company is defined as the total of the individual establishments under one ownership within an industry group or industry. Consequently, the same company may appear in several industry groups or industries if it has diversified activities. CONCENTRATION RATIOS AND THE SIC SYSTEM In essence, the SIC system seeks to establish spheres of economic activity which are unique and are distinguishable by the composite of similar characteristics which they have in common. These characteristics include similarity of products in terms of their uses, manufacturing processes, and materials used. The SIC system also attempts to bring together plants which specialize in making these products and account for a significant proportion of their total shipments, and plants which are economically significant in terms of their number, value added by manufacture, value of shipments, and number of employees. The standards also largely parallel the actual structure into which American businesses have grouped themselves, and it is the sum of these characteristics that make it the best single system for the largest number of uses. In the majority of instances, an industry is comprised of producers of similar goods or services. Usually, the products are made of similar materials and by similar processes, and the producers compete with one another. For the purpose of measuring concentration, the fact that the classification system is not based exclusively on the usage of the product is somewhat of a limitation, since in the economic concept of the market, it is immaterial whether products which are substitutable for each other are produced by the same processes, or are made from the same materials. In some cases, the industry and product definitions are too broad; that is, there are products that are included in the same category which do not serve the same function and are, thus, not substitutable for each other. For these cases, the concentration ratios for the industry may significantly understate the concentration in the individual product markets. In other cases, they are too narrow; that is, a single category fails to include products which are substitutable for example, metal and glass containers. For these cases, the concentration ratio may overstate the concentration in the market. Another limitation is the fact that the classification's industries and products are regularly being redefined. The need for redefinition arises particularly from the introduction of new products, the declining of older products, the introduction of new techniques, the growth of small fields into important industries, and similar dynamic developments. While necessary to keep the classification system abreast of the changing nature of the economy, an inevitable cost of redefinition is the loss of comparability for many categories over time. Another source of understatement of the concentration ratios is the method in which establishments are identified. The identification of which plants are controlled by which companies is based on census reports and available public records. As part of each census, each company is sent a questionnaire requiring them to list the plants which they operate directly or through their subsidiaries and affiliates. Any plant misreported as not being controlled by a company contributes to an understatement of the concentration of that company. A final limitation of the concentration ratios in the domestic market is due to the handling of imports and exports. Products imported and sold which require no further manufacturing, processing, or assembly are generally not included in the shipments for an industry, and tend to overstate the concentration in the market. Products manufactured in the United States and then exported are included in the shipments for an industry, and, therefore, tend to understate the concentration in the market. EXPLANATION FOR TABLES This table is based on the establishment approach described in the introduction. A "company" is defined as the total of the individual establishments under one ownership within an industry group or industry. Consequently, the same company may appear in several industry groups or industries if it has diversified activities. In addition, the following points should be kept in mind: 1. The top 4, 8, 20, and 50 companies are the largest for any particular year based on their value of shipments in that year and are, therefore, not necessarily the same companies from year to year. 2. For a few industries (indicated by footnotes), value added by manufacture rather than value of shipments has been used to determine the size of companies. This is considered necessary for industries in which the aggregated value of shipments contains a large amount of duplication due to large shipments of partially-finished products between establishments. An example is SIC 3312, Blast Furnaces and Steel Mills, in which there are extensive intra-industry sales and transfers of coke oven and blast furnace products to finishing plants. (The historical changes in the method for calculating value added by manufacture are described in the discussion of Tables 1 and 2.) 3. Ratios are shown for years prior to 1987 only if the SIC industry definitions were considered comparable. There were extensive revisions in the SIC system in 1987 and 1972 which, in Table 4, result in suppression of data for earlier years for a substantial number of industries. In general, the criteria used for judging the comparability of definitions was the effect of the change on the number of employees in the industry under the old and new definitions. If the SIC change caused more than a 2 percent change, the historical data were suppressed. Table 3 is new for 1987, and therefore, only the 1987 data are available. 4. The 1966, 1970, and 1976 concentration ratios are based on data developed in the Annual Survey of Manufacturers. Since the industry aggregates which provide the denominators in the concentration ratios are based on a sample, they are subject to sampling errors. Sampling errors for each industry, as well as a more complete description of the Annual Survey of Manufacturers sample, are shown in the appendix. HERFINDAHL-HIRSCHMANN INDEX The Herfindahl-Herschmann Index was shown for the first time in the 1982 Concentration Ratios in Manufacturing publication. The concept for this index was developed by Albert O. Hirschmann in 1945. in his book, National Power and Structure of Foreign Trade. A similar index was developed by Orris C. Herfindahl in his unpublished Ph.D. dissertation, "Concentration in the Steel Industry," at Columbia University in 1950. The Herfindahl-Hershmann Index is a truncated index and is calculated by squaring the concentration ratio for each of the top 50 companies or the entire universe (whichever is lower), and summing those squares to a cumulative total. The cumulative total is the Herfindahl-herschmann Index. This index is shown in table 3 at the two-digit industry group and all manufacturing-industries levels, and in table 4 at the four digit industry level. Application of the index is relatively straightforward. The higher the index, the more concentrated the industry group or industry is at the top. For example, consider an industry where the concentration ratio for the top 50 companies is 100. From the concentration ration, it is impossible to determine the amount accounted for by the top ranked companies. The Hefindahl-Herschmann Index would provide an insight to this determination. If each of the top 50 companies accounted for 2 percentage points of the 100 percent, the index would be 200. However, if the top ranked company accounted for 50 percentage points, and the remaining 49 companies slightly more than 1 percentage point each, the index would be 2550. CENSUS OF MANUFACTURES--CONCENTRATION RATIOS File Specifications CONTENTS General Data Specifications XX-9 File Layout XX-9 GENERAL DATA SPECIFICATIONS Data Fields. All files are recorded in dBase III+ format. Numeric data fields contain no alphabetic information. A few fields (those in percentage form) have explicit decimals, unlike corresponding data on tape, which carry only implied decimals. On the other hand, many data items are scaled in thousands or millions, and this is noted only in the file layouts. Any data field which is subject to any special conditions (e.g., suppression to avoid disclosure, not applicable) is preceded by an information flag field which contains explanatory codes. Standard Industrial Classification Titles. Records in the files are specific to a particular SIC classification, as identified by the SIC code. Titles for each manufacturing SIC are shown in the file MC87SIC.dbf on the CD-ROM. Labels for employment-size classes and percent estimated from administrative records are presented in the record layouts at the appropriate places, and they are also contained in the file MC87MISC.dbf on the CD-ROM. Titles for each of these labelled variables may be linked to data files using the dBase SET RELATION command, although only one relation can be defined at one time. EXTRACT's "Add Labels" function serves the same purpose. Index Files. Most of these data files have one or more index files that may be used to jump quickly to desired records (e.g., using the dBase FIND command) or to change the sequence of access. Available index files are listed in the abstract. Where noted in the abstract that an index is to the "first record for" a particular type of record, the index can be used during dBase FINDs, but the index must be shut off with SET INDEX TO in order to get access to records other than the first for each geographic area or SIC. FILE LAYOUT Most data items have an associated "flag" to describe any special conditions applicable. Flag 0 indicates normal data. Flag 6 indicates data greater than zero, but rounding to zero in the units displayed. Flag 7 identifies cells that have a relative standard error of 40 percent or more. Flags 1, 2, 4, 5, or 9 indicate the absence of data (e.g., due to disclosure). Flag 8 denotes the cumulation of more than one record into the presented summary. Data fields associated with flags 1 to 6 and 9 contain zeroes. In the geographic identification field in each file, NA fields contain either all zeroes or all 9's. See Geographic Code Structure. Census of Manufactures Concentration Ratios DATABASE STRUCTURE ----------------------------------------------------------------------------- Field Field Name Type Size Decimal Field Description ---------------------------------------------------------------------------- SIC C 4 0 1987 Standard Industrial Classification code YR C 2 0 Year (87,82,77,72,70,67,66,63,58,54,47,35) COMPANYF N 1 0 Flag for COMPANY COMPANY N 6 0 Number of companies SHIPMTSF N 1 0 Flag for SHIPMTS SHIPMTSE N 2 0 Relative standard error of SHIPMTS (1970 and 1966 only) SHIPMTS N 9 1 Value of shipments ($ million) LARGE4F N 1 0 Flag for LARGE4 LARGE4 N 3 0 Percent of shipments accounted for by 4 largest companies LARGE8F N 1 0 Flag for LARGE8 LARGE8 N 3 0 Percent of shipments accounted for by 8 largest companies LARGE20F N 1 0 Flag for LARGE20 LARGE20 N 3 0 Percent of shipments accounted for by 20 largest companies LARGE50F N 1 0 Flag for LARGE50 LARGE50 N 3 0 Percent of shipments accounted for by 50 largest companies HERFINDF N 1 0 Flag for HERFIND HERFIND N 4 0 Herfindahl-Hirschman Index for 50 largest companies SPECIALF N 1 0 Flag for SPECIAL SPECIAL N 3 0 Primary product specialization ratio (percent) COVERAGF N 1 0 Flag for COVERAG COVERAG N 3 0 Coverage ratio (percent) Record size 55 CENSUS OF MANUFACTURES--CONCENTRATION RATIOS User Notes This section will contain information relevant to the 1987 Census of Manufactures on CD-ROM, 1987: Concentration Ratios that indicates specific problems with the data, or that becomes available after the file is released. The cover letter to the updated information should be filed behind this page. User Notes will be sent to all users who (1) purchased their file (or technical documentation) from the Census Bureau and (2) returned the coupon