CHAPTER XXIX Analytical Report Exports from Manufacturing Establishments Abstract.......................................XXIX-3 Introduction...................................XXIX-5 Database Structure.............................XXIX-13 User Notes.....................................XXIX-17 ANALYTICAL REPORTS-EXPORTS FROM MANUFACTURING ESTABLISHMENTS Abstract CONTENTS Citation---------------------XXIX 3 Type of File-----------------XXIX 3 Universe Description---------XXIX 3 Subject-Matter Description---XXIX 3 Geographic Coverage----------XXIX 3 Technical Description--------XXIX 3 Reference Materials----------XXIX 4 Related Printed Reports------XXIX 4 Related Machine-Readable Data Files-----------------XXIX 4 Availability-----------------XXIX 4 CITATION Manufacturing Analytical Report Series on CD-ROM: Exports from Manufacturing Establishments, 1989 [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 these files is establishments with one or more paid employees primarily engaged in manufacturing. SUBJECT-MATTER DESCRIPTION: These data are exactly the same as shown in the printed Analytic Report Series report. The direct export data represent the actual export of manufactured goods as reported or estimated by manufacturers. As many establishments do not know if their products are exported, the directly reported numbers from the census are inflated upward to equal the official export data compiled by the Census Bureau's Foreign Trade Division. Supporting exports, i.e., the components, parts, supplies, etc. used by plants producing the export products, were estimated using the Bureau of Economic Analysis input-output accounts at the six-digit industry/commodity level of analysis. In addition to including employment in manufacturing establishments that supply materials and parts for the production of manufactured exports, the indirect estimates include employees in nonmanufacturing, such as farms and mines that provide materials input to manufacturing, plus various types of services required of nonmanufacturing industries in producing the manufactured goods to be exported, such as electricity, repair services, business services, transportation, utilities, etc. The indirect employment requirements could only be developed by an analytical approach which is not typical of other census reports. GEOGRAPHIC COVERAGE Data are shown for the U.S. and States. TECHNICAL DESCRIPTION: File Structure: dBASE III+ Most files also have dBASE ".NDX" index files and one or more label files that can be linked to them (e.g., with SIC titles). AREX1_89 Employment by Major Economic Sector by State (89, 88, 86) (tables 2a and 2b) 1. File sequence: by year (descending) by State 2. Available index: a. AREX1_89 By State by year AREX2_89 Shipments and Employment by 3-digit SIC by State (89, 88, 87) 1. File sequence: U.S. and States by SIC 2. Available indexes: a. AREX2S89: First record for each year within State b. AREX2Y89: First record for each State within year c. AREX2I89: by SIC by State by year 3. Linkable label files: a. MC87SIC (SIC titles) b. STCODE(State names) 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 Manufacturing Analytical Report Series: Exports from Manufacturing Establishments, 1989 (AR89-1). Includes the same statistics as this file for 1988 and 1989 (see reports for earlier years for 86 and 87 data). RELATED MACHINE-READABLE DATA FILES Manufacturing Analytical Report Series on CD-ROM: Selected Characteristics of Manufacturing Establishments That Export, 1987. Includes the number of establishments that export by "export intensity" (percent of goods produced that are exported) by 2- or 3- digit SIC for States and separately by employment-size class. (On CD-ROM 1E.) Census of Manufactures on CD-ROM, 1987: Geographic Area Series. Provides total shipments, employment and other statistics without regard to export activity for 4-digit SICs as well as the 2- and 3-digit SICs for the U.S., States, metropolitan areas, counties and places. (On CD-ROMs 1D and 1E). AVAILABILITY These files are included on Economic Census CD-ROM 1E (price $150). ANALYTICAL REPORTS-EXPORTS FROM MANUFACTURING ESTABLISHMENTS Introduction CONTENTS General Description----------XXIX 5 Relationship to Census Bureau's Series on Origin of Movement of U.S. Exports by State-----------XXIX 6 What's New in 1987-----------XXIX 6 Changes in the Data----------XXIX 6 Historical Information-------XXIX 7 Uses of the Data-------------XXIX 7 Detailed Description of Census Methods and Estimating Procedure-------XXIX 7 Indirect Requirements on Nonmanufacturing Industries----------------XXIX 10 Confidentiality of Data-----XXIX 12 GENERAL DESCRIPTION Exports From Manufacturing Establishments include estimates of the value of manufactured exports and export-related employment, both for directly exported goods and for the indirect requirements supporting manufactured goods. The estimates are shown by State, industry group (two-and three- digit SIC), and State by industry group. In addition, total value of shipments and total employment are presented so that data users may compare exports to overall manufacturing activity. The direct export data shown represent the actual export of manufactured goods by the manufacturers. In 1987, approximately $193.6 billion of manufactured goods were exported, and exports affected 1.2 million jobs in the manufacturing establishments. Information on direct export shipments, total value of shipments, and total employment are collected on the report forms for the 1987 Census of Manufactures. As many establishments do not know if their products are exported, the directly reported numbers from the census are inflated upward to equal the official export data compiled by the Census Bureau's Foreign Trade Division. The estimates of direct export shipments underestimate the importance of export trade to the domestic economy. For example, while the employment associated with automobile exports is shown in the transportation equipment industry data, the establishments supplying the auto industry with steel, windshields, transmissions, etc., would not report these sales as exports. Therefore, for the supplying industries, exports were more important than indicated by their direct export data. In order to put the importance of exports into proper perspective relative to the total export impact on all manufacturing and nonmanufacturing industries, the direct export values were analyzed using the latest Bureau of Economic Analysis (BEA) input-output (I-0) accounts at the six-digit industry/commodity level of detail. For 1987, the input- output analysis shows that the value of direct exports ($193.6 billion) requires an additional $185.2 billion of supporting shipments from other manufacturing establishments. (A breakdown of the total $378.8 billion by major group is shown in figure 1.) The impact on employment levels indicates an increase from the 1.2 million jobs estimated to be directly export related to a total of 2.8 million jobs directly and indirectly related to exports in manufacturing industries. In addition to including employment in manufacturing establishments that supply materials and parts for the production of manufactured exports, the indirect estimates include employees in nonmanufacturing, such as farms and mines that provide materials input to manufacturing, plus various types of services required of nonmanufacturing industries in producing the manufactured goods to be exported, such as electricity, repair services, business services, transportation, utilities, etc. The data also include estimates of employment associated with transporting the manufactured goods from plant-to-port and the employment of exporting establishments that purchase and resell the goods to foreigners or arrange their sale abroad. It should be emphasized that the employment estimates relate only to the export of manufactured goods and exclude employment related to the export of other nonmanufactured goods and services. They do not, therefore, measure the full employment effect of all U.S. exports. The indirect employment requirements could only be developed by an analytical approach which is not typical of other census reports. Although a number of economic assumptions believed to be reasonable were used, the estimates are not considered to be as precise as customary census statistics. The various assumptions and further comments on the quality of the estimates are included in the detailed discussion described later in this text. RELATIONSHIP TO CENSUS BUREAU'S SERIES ON ORIGIN OF MOVEMENT OF U.S. EXPORTS BY STATE In 1987, the Census Bureau began issuing two data series on the origin of exports. The two reports are different in several aspects. Each month the Census Bureau's Foreign Trade Division releases the Origin of Movement of U.S. Exports by State in its Summary of U.S. Export and Import Merchandise Trade (FT900) report. The data in this series include both manufactured and nonmanufactured products exported. Data are shown for the State where the merchandise began its export journey as reported by the exporter on the export declaration document filed with U.S. Customs at the port of export. This series provides data based on the "origin" of the commodity movement into international trade. That "origin" may be the production of site, and assembly point in a different State, or the port of export. The data are valued on a free alongside ship (f.a.s.) basis. The Census Bureau's Industry Division issues the annual report entitled Exports From Manufacturing Establishments. This includes data on export- related employment as well as the value of export shipments, but only for manufactured products that are exported. These data are shown for the State where the exported product was made as reported by manufacturing establishments in the Annual Survey of Manufactures or the Census of Manufactures. This series provides data based only on the production site as the "origin" of manufactured exports. In addition, the data are valued on a freight on board (f.o.b.) basis rather than and f.a.s. basis. WHAT'S NEW IN 1987 Reconciliation of the Undocumented Exports to Canada The estimates included reflect adjustments made to the benchmark trade data issued by our Foreign Trade Division. These adjustments primarily include the effect of the undocumented exports of manufactured commodities to Canada. The flow of goods traded between the United States and Canada is the largest in the world and growing at a rapid pace. However until 1990, discrepancies existed between the balance of trade published by the two countries. In reconciling the differences in the bilateral statistics, imports from each country were used as benchmark, and after adjustments were made for differences in definition of trade, valuation, timing, errors, and other known discrepancies, the residual difference was credited to undocumented or unreported exports. By 1987, these exports had reached such a level ($6.4 billion) as to prompt Statistics Canada and the Census Bureau to meet with the customs agencies of both countries to discuss ways of improving the accuracy of the trade statistics. The two statistical agencies set forth a target of replacing bilateral exports statistics with counterpart import statistics by 1990. As an interim step, the Census Bureau, with the release of the 1987 data, began adjusting the U.S. exports for undocumented shipments to Canada based on the Canada import data. Census Versus Sample The 1987 estimates are based on data collected in the census of manufactures which covers all establishments with one paid employee or more engaged in manufacturing as defined in the 1987 Standard Industrial Classification (SIC) Manual(1). During noncensus years, the estimates are compiled from the Annual Survey of Manufactures (ASM) which is a probability sample covering approximately 56,000 manufacturing establishments. CHANGES IN THE DATA The 1987 estimates are shown at the two-and three-digit industry level, and at the State by two-and three-digit levels. In the past, data were issued only at the two-digit major group level. Revision of The Industry Classification of Establishments In 1987, the Standard Industrial Classification (SIC) system was revised from the 1972 classification and its 1977 supplement. The result of this revision is that, at the aggregate level, some major groups comprise different mixes of establishments. Hence, comparisons between prior-year (1986) and current-year (1987) published totals should be viewed with caution. For 1987, this is particularly true for Major Groups 36, Electronic and Other Electric Equipment, and 38, Instruments and Related Products. HISTORICAL INFORMATION The survey was started in 1960 and initially measured only direct exports and direct export-related employment. In 1976, realizing that the relative importance of export trade was underestimated, the program was expanded to include analytically derived estimates for supporting exports and related employment. Prior to 1981, the program was conducted irregularly; the years of publication were 1960, 1963, 1966, 1969, 1971, 1972, 1976, 1977, 1980, and 1981. Since 1983, it has been conducted annually. USES OF THE DATA The exports form manufacturing establishments data, particularly the data on the number of export-related jobs, are used extensively in the analysis and promotion of exports by government agencies. The International Trade Administration (ITA) is the primary user of these data. Members of Congress have a particular interest in the data shown for their State. State agencies also are using the data to promote export activity. In addition, many consulting firms, academic institutions, and private manufacturing companies use the report to analyze trade, to make recommendations, and to compare export activity. DETAILED DESCRIPTION OF CENSUS METHODS AND ESTIMATING PROCEDURES The sections that follow describe the methodology used to develop the estimates of the effect of manufactured exports at the national level and by State. Directly Related Exports Manufacturers reporting in the Census of Manufactures were requested to "Report the value of products shipped for export. Include direct exports and products shipped to exporters or other wholesalers for export. Also include the value of products sold to the U.S. Government to be shipped to foreign governments. Do not include products shipped for further manufacture, assembly, or fabrication in the United States." Such directly reported exports understate the true value of all exports because many respondents do not know the final destination of the products produced in their plants. This included cases where the export was actually made by an independent exporting firm or where the actual shipments of the products are made from company-operated warehouses or other distribution points which did not allocate the export sales back to individual establishments. Estimated Value or Exports In order to provide a more accurate measure of the overall importance of exports industrially and geographically, the directly reported values have been adjusted to include estimates for the unreported exports, including exports of manufactured products by wholesalers, exporters, etc. The values for total exports of manufactured products are from official export data compiled by the Foreign Trade Division of the Bureau of the Census based on Shippers' Export Declarations. Because of this census measures only domestically manufactured exports, reexports are not included in the estimates. Merchandise exports were collected and compiled in accordance with the commodity classifications in Schedule B, Statistical Classification of Domestic and Foreign Commodities Exported from the United States. These classifications were rearranged into a structure related to the origin of production called the SIC-based export product classifications. This rearrangement was used by BEA to assign exports to six-digit I-O commodities. (The merchandise export data include exports from Puerto Rico and the Virgin Islands. To accurately estimate the effect of exports on the domestic economy, they must be subtracted since these areas are not included in the U.S. gross national product or the census. By the same token, exports from the United States to these territories must be added to the merchandise export data.) The export data are collected at port of export on a free alongside ship (f.a.s.) basis, thus they include freight and wholesale margins. To make the export data comparable to the data collected in the census, the margins were subtracted to derive a freight on board (f.o.b.) plant value. Six-digit I-O commodity margin rates from the 1977 benchmark I-O study were applied to these exports to derive the f.o.b. plant value of exports. The merchandise export data are collected on an individual commodity transaction basis, while the export data in the census are collected on the basis of all commodities exported by the industry. The merchandise export data were converted to an industry basis using 1987 Census of Manufactures data on the distribution of product shipments by industry. The difference between these National estimates of export shipments and reported export shipments of manufacturing establishments were then allocated to geographic areas (see Allocation of Unreported Value of Exports) and added to the reported data to develop estimates of total f.o.b. value of exports of manufacturing establishments by State. Allocation of Unreported Value of Exports Differences between the directly reported exports and the National export estimates f.o.b. plant have been allocated at three-digit SIC levels to States. The estimating procedure assumes that export shipments not reported by manufacturing establishments are directly related to the geographic distribution of the total employment of the industry and the export shipments reported for specific States. There are no measures available regarding the possible range of error associated with this estimating technique. To the extent that directly reported exports are a large proportion of the total exports, the geographic distribution of the total will be reliable. However, where the directly reported exports are a relatively small portion of total exports, the geographic distribution may be subject to considerable error if the basic assumption is not valid for a specific industry. Estimating Direct Export-Related Employment Employment related to direct exports was not collected in the census. However, the total value of shipments, as well as total employment, was collected for each establishment. Employment related to direct exports was then calculated for each establishment by multiplying the total employment by the ratio of export shipments to total shipments. These calculated export employment figures were then aggregated by industry and area to produce the directly reported export employment estimates that appear in the report. Since this estimate has the same weakness in underreporting as the export-shipment, it was necessary to estimate total export employment by a procedure similar to the one employed in deriving export shipments. The Meaning of Indirect Requirements The indirect requirements refers to all of the inputs acquired by the manufacturing establishments to produce the product being exported and the requirements of these suppliers, etc. As an example, the export value of engines and turbines in 1987 amounted to $3.1 billion as goods left the manufacturing plant. To produce the $3.1 billion of engines and turbines, it was necessary for these plants to purchase steel and additional metals from other plants, various other materials, containers for shipping, electric energy from utility firms, oil and lubricants for maintenance, a variety of business and professional services, etc. Producers of the various inputs and services which are sold to engine and turbine producers must also purchase inputs and various services from farms, mines, factories, utilities, transportation, etc. These secondary suppliers must, in turn, purchase inputs from their suppliers. At the aggregate level, this lengthy chain of transactions leading to the production of engines and turbines is referred to as the indirect requirements for engines and turbines (or for any other final products). Total indirect requirements are derived by applying BEA's input-output total requirements tables to the products for which a solution is sought. Estimating Indirect Requirements for Manufactured Exports 1. The exports by commodity (a bill of goods) are deflated to 1977 prices using current year weighted product prices. 2. The total industry domestic output required to supply these exports is estimated using the I-O total requirements coefficients tables. 3. Industry output is reflated back to current dollars. The estimate of domestic output requires that the import content of industry requirements be factored out of the estimates. For example, suppose that engine and turbine manufacturers purchased $200 million of steel as inputs, some of the steel purchased was imported, which has no effect on domestic output or employment, thus must be excluded. The import content of manufactured exports was estimated by the following process: 1. The commodity-by-commodity total requirements table was multiplied times the export bill of goods to estimate total commodity requirements. 2. Exports were subtracted to obtain only indirect requirements. 3. Total indirect requirements were multiplied by an import content ratio to get the import content of the commodity output required for exports. 4. The import content ratios were derived by BEA using the following formula: IM Import share = _____________________ S + /\IN X+IM+ETAX Where:IM = Imports assigned to commodities are from census merchandise import data adjusted for U.S. territories and assigned to I-O commodities by BEA. S = Product shipments are obtained from the Census of Manufactures. /\IN = Change in inventories during year from the Census of Manufactures. X = Exports assigned to commodities are from census merchandise export data adjusted for U.S. territories and assigned to I-O commodities by BEA. ETAX = Exercise taxes. The estimated imports by commodity was subtracted from the exports by commodity. This resulted in a bill of goods which excludes import requirements. This adjusted bill of goods was multiplied by the industry- by commodity total requirements table to give the estimated domestic industry output required to meet the 1987 level of manufactured exports. Import content ratios for all manufacturing commodities were derived using 1987 data. For all other commodities, data from the latest (1986) annual I-O accounts were used to calculate the ratios. The indirect requirements were converted from an I/O to a three-digit SIC basis for allocation to States (see below). It should be noted that indirect requirements include the demands of an industry upon itself. Thus, the producer of a turbine may purchase components from another plant in the engine and turbine industry, or a cold rolled steel plant may purchase hot rolled steel from another plant in primary iron and steel manufacturing. Allocating Indirect Requirements of Manufacturing Industries by State To describe this estimation process, a part of the actual work sheet for SIC Industry 201, Meat Products, is being reproduced (figure 5). To simplify, only one State is shown, but the same process was applied to the other States. In lines A, B, and C, a ratio of about 1 percent is obtained for the U.S. industry; that is, after subtracting direct exports for the entire industry, about 1 percent of the supply is required to support further exports. On line D, the total shipments for the State in 1987 is shown. On the next line, the direct exports from the State is given, and line F shows the available supply for domestic uses produced in the State. The industry ratio of indirect requirements to total supply is applied to the State on line I. This assumes that each State contributes an equal proportion of its domestic supply to the product of finished manufactured goods to be exported. Figure 5. Estimating Indirect Requirements in Manufacturing by State, SIC Industry 201, Meat Products (Money figures in million dollars, except where noted) U.S. industry ration Estimates A. Indirect requirements 979.6 B. Total value of shipments (1987 census) less direct exports 73 658.6 C. Ratio A/B .013299 State calculations Alabama D. Value of shipments 1 169.8 E. Direct exports 130.8 F. Domestic supply (D-E) 1 039.0 G. Employment (1,000) 12.8 H. Shipments per employee (D/G) (1,000) 91.4 I. Indirect exports (FxC) 13.8 J. Indirect employment in manufacturing (I/H) (1,000) 0.1 In the remaining lines of the figure, the values of goods are converted to employment. For each State, the value of shipments per employee is obtained from 1987 Census of Manufacturers data (line H). This value is then divided into the indirect requirements to arrive at the number of employees involved in production of goods for exports (line J). The total indirect employment in manufacturing for each State is obtained by summing the employment in each manufacturing industry. Estimating Export-Related Employment for Auxiliaries Large firms in manufacturing often have central administrative offices, warehouses, garages, repair shops, research centers, etc., which support or provide services to the manufacturing plants. These establishments, usually referred to as auxiliaries, also contribute indirectly to the production of merchandise for export. A special tabulation of the employment of auxiliaries by manufacturing industry by State was prepared and estimates made of the proportion of employment attributable to exports. The procedure provided an estimate of 204,900 employees as export-related in 1987. The export-related employment estimated for auxiliaries was based on the ratio of the total export-related manufacturing employment at the two- digit National level to the total manufacturing employment at the two-digit National level (see EXPPCTE in AREX2_..). The numerator of this ratio includes the estimates for both direct and indirect export-related manufacturing employment. Each of the ratios was applied to the auxiliaries employment at the corresponding two-digit level within a State and resulting figures summed to a total for the State. For example, assume State X includes only two different industry groups as shown below: Auxiliary ratio Industry Industry group 25 group 39 1. Total auxiliary employment 3 000 10 000 2. Two-digit National ratio from table 4, column J .020 .075 3. Export-related auxiliary employment (1x2) 60 750 Thus, the total export-related employment for auxiliaries for State X is the sum of the two figures in line 3 above, or 810. INDIRECT REQUIREMENTS ON NONMANUFACTURING INDUSTRIES The estimates of output requirements on nonmanufacturing industries as inputs to manufactured goods for exports are used to estimate the employment in these sectors required to support manufactured exports. The estimates were made using the 1977 I-O accounts data on employment by I-O industry. Indirect industry output was multiplied by the 1977 ratio of employment to industry output to derive industry employment. The 1977 data were used since they are consistent with the I-O accounts; using these data assumes no change in productivity in these sectors since 1977. Following is a description of how estimates measuring employment required to move the manufactured products from factory to port, as well as employment of export wholesalers, were made. Plant-to-Port Indirect Requirements The port value of manufactured exports in 1987 amounted to approximately $221.5 billion, and it was estimated that the f.o.b. plant value of these goods amounted to $193.6 billion. The difference of $279 billion represents transportation costs of moving these goods, mainly to seaports but also to airports or rail and truck border crossing points. In addition, there is the margin, or markup, of exporters, which is a component of wholesale trade. Using rates from the 1977 I-O account, margins were estimated and allocated as follows: $5.4 billion were attributed to transportation and $22.6 billion to wholesale trade. These figures were then included in the I-O computation. Allocating Nonmanufacturing Employment by State The statistical procedures described above resulted in National estimates of persons engaged in nonmanufacturing industries induced by U.S. manufacturing exports. These are industries supplying materials to manufacturing for further processing, and services and supplies of all kinds to manufacturing industries. The second phase of this procedure was to allocate these persons-engaged estimates by State. Procedures used for this purpose were the following: For agriculture, where about 145,800 persons were estimated to be supplying food and fiber for processing to manufactured goods which were later exported, the State distribution, as reported in the 1987 Current Population Survey, was used. For mining, with only 64,400 persons engaged to be allocated, the State distribution of employment, as reported in the 1987 Current Population Survey, was also used. There were 1,207,000 persons engaged in trade (including the plant-to- port exporters) estimated to be supporting production of manufactured export goods or supporting the plant-to-port exporters. These include persons purchasing agricultural goods, steel, ores, and various other materials and supplies form producers and selling them to manufacturers for further processing into goods which were exported. Based upon review of more detailed I-O data, 75 percent of this employment was assigned to trade and 25 percent to plant-to-port exporters. The wholesale trade export employment was allocated to States based upon the distribution of wholesale trade in the 1987 Current Population Survey. The allocation of plant-to port exporters was based on a different pattern as this category includes firms purchasing goods from manufacturers and exporting them; that is, selling them to foreign firms or governments. Here it was decided to use the 1987 Census of Wholesale Trade, because within the Census, merchant wholesalers are further classified to distinguish those primarily engaged in exporting. A wholesale establishment exporting 50 percent or more of sales in classified as a merchant wholesale exporter, and those exporting less than 50 percent of sales are "plain" merchant wholesalers. For 1987, establishments with approximately 51,000 employees were classified as "primary exporters" from the 4.5 million employees in merchant wholesale. The State allocation of these "primary export" wholesalers is markedly different from other wholesalers with heavy concentration in New York, California, and Florida. It was, therefore, concluded that these data for export wholesalers should be given heavy weight in allocating the persons in wholesale establishments estimated as being involved in the export of manufactured goods. Accordingly, the plant-to-port wholesale employment was allocated by State on the basis of the 1987 Census of Wholesale Trade "primary exporters" industry. The remaining nonmanufacturing employment (mostly in transportation, utilities, and business services) was judged to be closely allied in terms of location of the State distribution of manufacturing; that is, they supply services in the sense of satellites of manufacturing plants in the vicinity. This employment was, therefore, allocated by State proportionately based on the State distribution of export-related manufacturing employment. Quality of the Estimates As previously stated, many of the statistical processing steps, beyond the collection of the data on the direct value of exports, involved a series of assumptions so that the final estimates of indirect export- related employment by State are not of the usual Census quality; that is, their estimates go beyond the customary Bureau of the Census function of collecting and compiling data in a carefully designed survey. This estimates included in reports prior to 1976 of the State distribution of jobs directly related to exports were criticized as being incomplete because of the omission of indirect employment, and the Bureau was urged to expand the estimated to measure the full impact by State. The task could be undertaken only because of the availability of BEA's input-output accounts which show interindustry relations and transactions between producing industries and final purchases. It is, however, a matrix based on estimated relationships between millions of buyers and sellers and is, therefore, subject to estimation error. Flaws in the estimated relationships might carry over to this project. Furthermore, relationships estimated or 1977 may not apply for subsequent years. However, it should be recognized that the I-O matrix is compiled with painstaking caution by highly skilled technicians, and further that the interlocking system provides careful cross checks of itself. At a number of later stages in the process, assumptions were required on how various estimates of allocations were to be made. These assumptions were carefully weighed and methods developed which took account of quality requirements and feasibility. For example, suppose a manufactured raw material was made in 20 States, and 10 percent of the domestic sales went to producers of finished products for export. It was not feasible, or even possible, to estimate specifically how much of the raw material from each State found its way into exported manufactured goods, so each State was assumed to have 10 percent of its domestic sales going to exports indirectly. One might even argue that, for a given State, even if it could be determined that 12 percent was actually going to exports indirectly in a physical sense, that for economic purposes 10 percent was preferable because the manufacturer of the finished product would be indifferent as to where the supply came from. In any case, if there are errors due to the assumption of proportionality, there are considered to be relatively minor. In the Bureau of the Census surveys of the manufacturing sector, working proprietors and partners are omitted by convention and are, therefore, omitted from these export-related job estimates. They are basically omitted because they are judged to be a negligible volume. That is also the case here. Furthermore, the procedure for converting the estimates for the effect of manufactured indirect inputs is judged to underestimate employment by about 40,000 to 50,000. This is because shipments of parts suppliers were deflated to obtain employment by an aggregate which sometimes included finished goods producers in the same State and industry. The parts producers tend to require more employment per dollar of shipments than the finished goods producing firm in the same industry. Since the conversion ratio for the parts shipments was based upon aggregates for both types of plants, the resulting estimate of the export-related employment is a bit low. All three of the comments described above underestimate export-related employment. Working in the opposite direction are advances in productivity (more output per unit of labor input). The I-O data used were for 1977, the latest benchmark data available. For this problem, the relevant area of concern is the nonmanufacturing industries where an important step was conversion of the value of indirect requirements to compensation of labor by use of the 1977 I-O rations. While nonmanufacturing industries are the most difficult to measure for productivity advances, they are surely positive overall from 1977 to 1987 and, therefore, result in too high an estimate of the overstatement in number of jobs is not feasible. Finally, not part of the basic work but worthy of comment are the percentages calculated form the civilian employment and private-sector employment shown in table 2. The export-related jobs are estimates of the States in which the jobs are located. The civilian and private-sector employment come from the Census Bureau's Current Population Survey (CPS). Because this is a household survey, the State aggregates are based on where the people live. An attempt was made in earlier years to adjust the CPS household data by shifting estimates from a "where-live" basis to a "where-work" basis, but the resulting percentages turned out to be so slightly changed that it did not seem necessary to make the adjustments. CONFIDENTIALITY OF DATA The Bureau of the Census is prohibited by law from publishing any statistics that disclose information reported by individual establishments. Whenever the total value of shipments and/or exports shipments cannot be shown without disclosing information for individual establishments, the complete line has been suppressed. Data for these suppressed lines are, however, included in higher level totals. Additional disclosure analysis is performed for exports and export-related employment that can be suppressed even though value of shipment data are publishable. _____________ 1. Standard Industrial Classification Manual: 1987: For sale by Superintendent of Documents, U.S. Government Printing Office, Washington, DC 20402. Stock No. 041-001-00314-2. ANALYTICAL REPORTS-EXPORTS FROM MANUFACTURING ESTABLISHMENTS on CD-ROM DATABASE STRUCTURE File AREX1_89: Table 2 Sectors for States --------------------------------------------------------------------------- Field Field Name Type Size Decimal Field Description --------------------------------------------------------------------------- YR C 2 0 Year ST C 2 0 FIPS State Code TEXT C 22 0 State name EMPLCIV N 8 1 Civilian employment (1,000) EMPLPRIV N 8 1 Prive sector employment (1,000) EMPLEXP N 6 1 Manufacturing industries employment related to exports (1,000) EMPLTRAD N 6 1 Trade employment related to manufactured exports (1,000) EMPLSVCS N 6 1 Business services employment related to mfrd exports (1,000) EMPLTRAN N 6 1 Trans., communications, & util, empl. related to exports (1K) EMPLOTHR N 6 1 Mining, agr., & othrer employment related to mfrd exports (1k) EMPLNONM N 6 1 Total nonmfg industries empl. related to mfrd exports (1,000) EMPLTOT N 6 1 Total employment related to manufactured exports (1,000) EXPCIV N 4 1 Export-related employment as % of civilian employment (%) EXPPRIV N 4 1 Export-related employment as % of private sector employment (%) Record size 93 ANALYTICAL REPORTS-EXPORTS FROM MANUFACTURING ESTABLISHMENTS on CD-ROM DATABASE STRUCTURE File AREX2_89: Table 3-5 States by SIC --------------------------------------------------------------------------- Field Field Name Type Size Decimal Field Description --------------------------------------------------------------------------- YR C 2 0 Year ST C 2 0 FIPS State Code SIC C 4 0 1987 Standard Industrial Classification code SICTEXT C 47 0 SIC title SHIPMTSF N 1 0 Flag for SHIPMTS SHIPMTS N 9 1 Value of shipments ($ million) SHIPDIRF N 1 0 Flag for SHIPDIR SHIPDIRE N 2 0 Relative standard error for SHIPDIR (N/A for 1987) SHIPDIR N 8 1 Value of direct exports shipments ($ million) SHIPSUPF N 1 0 Flag for SHIPSUP SHIPSUP N 8 1 Value of supporting exports shipments ($ million) SHIPEXPF N 1 0 Flag for SHIPEXP SHIPEXP N 8 1 Value of total export-related shipments ($million) EXPPCTSF N 1 0 Flag for EXPPCTS EXPPCTS N 5 1 Total export-related as a percent of total shipments (%) EMPLOYEF N 1 0 Flag for EMPLOYE EMPLOYE N 7 1 Manufacturing employment, total EMPLDIRF N 1 0 Flag for EMPLDIR EMPLDIRE N 2 0 Relative standard error for EMPLDIR (N/A for 1987) EMPLDIR N 6 1 Manufacturing employment related to direct export shipments EMPLSUPF N 1 0 Flag for EMPLSUP EMPLSUP N 6 1 Manufacturing employment related to supporting export shipment EMPLEXPF N 1 0 Flag for EMPLEXP EMPLEXP N 6 1 Manufacturing employment related to toal export shipments EXPPCTEF N 1 0 Flag for EXPPCTE EXPPCTE N 5 1 Total export-related as a percent of total employment (%) Record size 138 ANALYTICAL REPORTS-EXPORTS FROM MANUFACTURING ESTABLISHMENTS User Notes This section will contain information relevant to the Exports from Manufacturing Establishments report 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 following the CD-ROM File Information.