Geographic Area Statistics: 2001

U.S. Department of CommerceEconomics and Statistics AdministrationU.S. CENSUS BUREAU

Annual Survey of Manufactures

Issued January 2003

M01(AS)-3

U S C E N S U S B U R E A UHelping You Make Informed Decisions

The staff of the Manufacturing and Con-struction Division prepared this report.Judy M. Dodds, Assistant Chief for Cen-sus and Related Programs, was respon-sible for the overall planning, manage-ment, and coordination. Robert Reinard,Chief, Consumer Goods Industries Branch,assisted by Suzanne Conard, KayleneHanks, and Jim Jamski, Section Chiefs,Kenneth Hansen, Chief, InvestmentGoods Industries Branch, assisted by MikeBrown, Chris Savage, and MilbrenThomas, Section Chiefs, NathanielShelton, Chief, Primary Goods IndustriesBranch, assisted by Nancy Higgins, PaulMarck, and Joanna Nguyen, SectionChiefs, and Robert Rosati and TomFlood, Special Assistants, performed theplanning and implementation. TimothyAlthouse, Stephanie Angel, BillBaldwin, Melissa Berry, ChrisBlackburn, Luis Blanco, LarryBlumberg, Phillip Brown, BrendaCampbell, Theresa Crowley, ChrisCunningham, Vance Davis, TracyFenwick, Marquita Flemmings,Melinda Ford, Allen Foreman, KellieFriedrich, Matthew Harris, VeraHarris-Bourne, Karen Harshbarger,James Hinckley, Richard Hough,Walter Hunter, Tom Ickes, DeborahJohnson, Marvella Jones, EvelynJordan, Cathy Knudsen, Robert Lee,Jennifer Leotta, John Linehan, KeithMcKenzie, Philippe Morris, BettyPannell, Bridgett Parker-Bell, GloriaPeebles-Butler, Michael Perkinson,Deanna Pickerall, Jazmin Rose,LaTanya Steele, Aronda Stovall, SueSundermann, AthanasiosTheodoropoulos, Dora Thomas,Ronanne Vinson, Keeley Voor,Denneth Wallace, Danielle Welch,Tempie Whittington, and Ernest Wilsonprovided primary staff assistance.

Paul Hsen, Assistant Chief for Researchand Methodology Programs, assisted byStacey Cole, Chief, Manufacturing Pro-grams Methodology Branch, and RobertStruble, Section Chief, provided the math-ematical and statistical techniques as wellas the coverage operations. JeffreyDalzell and Cathy Gregor provided pri-mary staff assistance.

Mendel D. Gayle, Chief, Forms, Publica-tions, and Customer Services Branch,assisted by Kimberly DePhillip andBaruti Taylor, Section Chiefs, performedoverall coordination of the publication pro-cess. Patrick Duck, Chip Murph, WandaSledd, and Veronica White provided pri-mary staff assistance.

The Economic Planning and CoordinationDivision, Lawrence A. Blum, AssistantChief for Collection Activities andBeverly M. Eng, Assistant Chief for Post-Collection Current Activities, assisted byDennis Shoemaker, Chief, AnalyticalBranch, and Brandy Yarbrough, Chief,Edit Branch, and Sheila Proudfoot,Richard Williamson, and Andrew W.Hait were responsible for developing thesystems and procedures for data collec-tion, editing, review, correction and dis-semination.

The staff of the National Processing Center,Judith N. Petty, Chief, performed mailoutpreparation and receipt operations, clericaland analytical review activities, data key-ing, and geocoding review.

The Geography Division staff developedgeographic coding procedures and associ-ated computer programs.

The Economic Statistical Methods and Pro-gramming Division, Howard Hogan,Chief, developed and coordinated the com-puter processing systems. BarrySessamen, Assistant Chief for Post Col-lection, assisted by Gary T. Sheridan,Chief, Macro Analytical Branch, EdwardJohnson, Section Chief, supervised thepreparation of the computer programs andwas responsible for the design and imple-mentation of the computer systems.

Computer Services Division, KennethRiccini, Chief, performed the computerprocessing.

Kim D. Ottenstein, Cynthia G. Brooks,and Crystal M. Pate of the Administrativeand Customer Services Division, Walter C.Odom, Chief, provided publications andprinting management, graphics design andcomposition, and editorial review for printand electronic media. General directionand production management were pro-vided by James R. Clark, Assistant Divi-sion Chief, and Gary J. Lauffer, Chief,Publications Services Branch.

ACKNOWLEDGMENTS

Annual Survey of Manufactures

Geographic Area Statistics: 2001 Issued January 2003

M01(AS)-3

U.S. Department of CommerceDonald L. Evans,

Secretary

Samuel W. Bodman,Deputy Secretary

Economics andStatistics Administration

Kathleen B. Cooper,Under Secretary for

Economic Affairs

U.S. CENSUS BUREAUCharles Louis Kincannon,

Director

Nancy M. Gordon,Acting Principal AssociateDirector for Programs

Frederick T. Knickerbocker,Associate Directorfor Economic Programs

Thomas L. Mesenbourg,Assistant Directorfor Economic Programs

William G. Bostic, Jr.,Chief, Manufacturingand Construction Division

ECONOMICS

AND STATISTICS

ADMINISTRATION

Economicsand StatisticsAdministration

Kathleen B. Cooper,Under Secretaryfor Economic Affairs

U.S. CENSUS BUREAU

Charles Louis Kincannon,Director

Hermann Habermann,Deputy Director andChief Operating Officer

Introduction

SCOPE

The Annual Survey of Manufactures (ASM) provides sampleestimates of statistics for all manufacturing establish-ments with one or more paid employee. Manufacturing isdefined as the mechanical, physical, or chemical transfor-mation of materials or substances into new products. Theassembly of components into new products is also consid-ered manufacturing, except when it is appropriately classi-fied as construction.

Establishments in the manufacturing sector are oftendescribed as plants, factories, or mills and typically usepower-driven machines and materials-handling equipment.Also included in the manufacturing sector are some estab-lishments that make products by hand, like custom tailorsand the makers of custom draperies. While manufacturerstypically do not sell to the public, some establishmentslike bakeries and candy stores that make products on thepremises may be included.

GENERAL

This report presents manufacturing establishment statis-tics from the 2000 Annual Survey of Manufactures (ASM).Three reports are issued from this survey.

The report Statistics for Industry Groups and Industriespresents 2000 and earlier years data at the three-, four-,five-, and six-digit North American Industry ClassificationSystem (NAICS) levels. The 1997 data are from the 1997Economic Census and include revisions made to the datasince its initial release. The report also includes a historictable with data for the all-manufacturing level back to1977. This report includes statistics for employment, pay-roll, value added by manufacture, cost of materials con-sumed, value of shipments, detailed capital expenditures,supplemental labor costs, fuels and electric energy used,and inventories by stage of fabrication.

The report Geographic Area Statistics presents similardata for each state and the District of Columbia. Thisreport includes statistics at the NAICS three- and four-digitlevels for employment, payroll, value added by manufac-ture, cost of materials consumed, value of shipments, andtotal capital expenditures. Data for detailed capital expen-ditures, supplemental labor costs, fuels and electricenergy used, and inventories by stage of fabrication arepresented at the state level.

The report Value of Product Shipments presents shipmentsdata for the 473 six-digit NAICS product groups andapproximately 1,500 seven-digit NAICS product classes.

COMPARABILITY WITH HISTORIC DATA

The adoption of the North American Industry Classifica-tion System (NAICS) in the 1997 Economic Census has hada major impact on the comparability of current and his-toric data. Approximately half of the industries in themanufacturing sector of NAICS do not have comparableindustries in the Standard Industrial Classification (SIC)system that was used in the past.

While most of the change affecting the manufacturing sec-tor was change within the sector, some industries leftmanufacturing and others came into manufacturing.Prominent among those that left manufacturing are log-ging and portions of publishing. Prominent among theindustries that came into the manufacturing sector arebakeries, candy stores where candy is made on the pre-mises, custom tailors, makers of custom draperies, andtire retreading. The net effect of the classification changesare such that if the 1997 value of shipments data for allmanufacturers were tabulated on an SIC basis, it would beapproximately 3 percent higher.

Another change resulting from the conversion to NAICS isthat data for auxiliaries and central administrative offices(CAOs) associated with manufacturers are not presentedwith the manufacturing data.

It should also be noted that while the Census Bureau pub-lished ‘‘new capital expenditures’’ in the past, starting with1997 it is publishing ‘‘total capital expenditures.’’ The his-toric data presented in this report have been adjusted tobe consistent with the new policy.

DISCLOSURE

In accordance with federal law governing census reports(Title 13, United States Code), no data are published thatwould disclose the operations of an individual company.

The disclosure analysis for the industry statistics files isbased on the total value of shipments. When the totalvalue of shipments cannot be shown without disclosinginformation for individual companies, the complete line issuppressed except for capital expenditures. However, the

Introduction vAnnual Survey of Manufactures

U.S. Census Bureau

suppressed data are included in higher-level totals. A sepa-rate disclosure analysis is performed for capital expendi-tures that can be suppressed even though value of ship-ments data are published.

AVAILABILITY OF MORE FREQUENTMANUFACTURING DATA

The Census Bureau conducts the ASM in each of the4 years between the economic census, which is collectedfor years ending in 2 and 7. The economic census – manu-facturing is the sample frame from which the ASM ischosen and presents more detailed data than the ASM.

In addition, the Census Bureau conducts a Current Indus-trial Reports (CIR) program. The CIR publishes product sta-tistics for selected manufacturing industries at the U.S.level annually and, in some cases, monthly and/or quar-terly.

The Census Bureau also conducts the monthly Manufactur-ers’ Shipments, Inventories, and Orders (M3) program,which publishes detailed statistics for manufacturingindustries at the U.S. level.

INDUSTRY CLASSIFICATION OF ESTABLISHMENTS

Each of the establishments canvassed in the ASM was clas-sified in 1 of 473 manufacturing industries in accordancewith the industry definitions in the 1997 NAICS Manual.

In the NAICS system, an industry is generally defined as agroup of establishments that have similar production pro-cesses. To the extent practical, the system uses supply-based or production-oriented concepts in defining indus-tries. The resulting group of establishments must besignificant in terms of number, value added by manufac-ture, value of shipments, and number of employees.

The coding system works in such a way that the defini-tions progressively become narrower with successiveadditions of numerical digits. In the manufacturing sector,there are 21 subsectors (three-digit NAICS), 86 industrygroups (four-digit NAICS), 184 NAICS industries (five-digitNAICS) that are comparable with Canadian and Mexicanclassification, and 473 U.S. industries (six-digit NAICS).Product classes and products of the manufacturing indus-tries have been assigned codes based on the industryfrom which they primarily originate. In the NAICS system,there are about 1,500 product classes (seven-digit codes),about 6,000 census products, and an additional 3,700 CIRproducts (ten-digit codes). The ten-digit products are con-sidered the primary products of the industry with thesame first six digits.

In the 1997 Economic Census – Manufacturing, all estab-lishments were classified in particular industries based onthe products they produced. If an establishment made

products of more than one industry, it was classified in theindustry with the largest product value. In the censusyears, there were no resistance rules or frozen industries.

In ASM years, establishments included in the ASM samplewith certainty weights are reclassified by industry only ifthe change in the primary activity from the prior year issignificant or if the change has occurred for two succes-sive years. This procedure prevents reclassification whenthere are minor shifts in product mix.

In ASM years, establishments included in the ASM samplewith noncertainty weight are not shifted from one industryclassification to another. They are retained in the industrywhere they were classified in the base census year. How-ever, in the following census year, these ASM plants areallowed to shift from one industry to another.

The results of these rules covering the switching of plantsfrom one industry classification to another are that someindustries comprise different mixes of establishments indifferent survey years. Hence, comparisons between prior-year and current-year published totals, particularly at thesix-digit NAICS level, should be viewed with caution. Thisis particularly true for the comparison between the datashown for a census year versus the data shown for theprevious ASM year.

ESTABLISHMENT BASIS OF REPORTING

The ASM is conducted on an establishment basis. A com-pany operating at more than one location is required tofile a separate report for each location or establishmentselected in the sample. Companies engaged in distinctlydifferent lines of activity at one location are requested tosubmit separate reports if the plant records permit such aseparation and if the activities are substantial in size.

DUPLICATION IN COST OF MATERIALS AND VALUEOF SHIPMENTS

Data for cost of materials and value of shipments includevarying amounts of duplication, especially at higher levelsof aggregation. This is because the products of one estab-lishment may be the materials of another. The value addedstatistics avoid this duplication and are, for most pur-poses, the best measure for comparing the relative eco-nomic importance of industries and geographic areas.

VALUE OF INDUSTRY SHIPMENTS COMPARED WITHVALUE OF PRODUCT SHIPMENTS

The 1997 Economic Census – Manufacturing shows valueof shipments data for industries and products. In theindustry statistics tables and files, these data representthe total value of shipments of all establishments classi-fied in a particular industry. The data include the ship-ments of the products classified in the industry (primaryto the industry), products classified in other industries

vi Introduction Annual Survey of Manufactures

U.S. Census Bureau

(secondary to the industry), and miscellaneous receipts(repair work, sale of scrap, research and development,installation receipts, and resales). Value of product ship-ments shown in the products statistics tables and filesrepresent the total value of all products shipped that areclassified as primary to an industry regardless of the clas-sification of the producing establishment.

DOLLAR VALUES

All dollar values presented are expressed in current dol-lars. Consequently, when making comparisons with prioryears, users of the data should consider the changes inprices that have occurred.

All dollar values are shown in thousands of dollars.

ABBREVIATIONS AND SYMBOLS

The following abbreviations and symbols are used withthe data:

A Standard error of 100 percent or more.

D Withheld to avoid disclosing data of individualcompanies; data are included in higher leveltotals.

N Not available or not comparable.

S Withheld because estimates did not meet pub-lication standards.

X Not applicable.

r Revised.

nec Not elsewhere classified.

nsk Not specified by kind.

– Represents zero (page image/print only).

NAICS North American Industry Classification Sys-tem.

Introduction viiAnnual Survey of Manufactures

U.S. Census Bureau

Appendix A.Explanation of Terms

BEGINNING- AND END-OF-YEAR INVENTORIES

Respondents were asked to report their beginning-of-yearand end-of-year inventories at cost or market. Effectivewith the 1982 Economic Census, this change to a uniforminstruction for reporting inventories was introduced for allsector reports. Prior to 1982, respondents were permittedto value inventories using any generally accepted account-ing method (FIFO, LIFO, market, to name a few). Beginningin 1982, LIFO users were asked to first report inventoryvalues prior to the LIFO adjustment and then to report theLIFO reserve and the LIFO value after adjustment for thereserve.

Inventory Data by Stage of Fabrication

Total inventories and three detailed components (finishedgoods, work-in-process, and materials, supplies, fuels,etc.) were collected.

When using inventory data by stage of fabrication, itshould be noted that an item treated as a finished productby an establishment in one industry may be reported as araw material by an establishment in a different industry.For example, the finished-product inventories of a steelmill would be reported as raw materials by a stampingplant. Such differences are present in the inventory figuresby stage of fabrication shown for all publication levels.

TOTAL CAPITAL EXPENDITURES (NEW AND USED)

For establishments in operation and any known plantsunder construction, manufacturers were asked to reporttheir new and used expenditures for (1) permanent addi-tions and major alterations to manufacturing establish-ments and (2) machinery and equipment used for replace-ment and additions to plant capacity if they were of thetype for which depreciation accounts were ordinarily main-tained.

Totals for expenditures include the costs of assets leasedfrom nonmanufacturing concerns through capital leases.New facilities owned by the federal government but oper-ated under contract by private companies and plant andequipment furnished to the manufacturer by communitiesand nonprofit organizations are excluded. Also excludedare expenditures for land and cost of maintenance andrepairs charged as current operating expenses.

For any equipment or structure transferred for the use ofthe reporting establishment by the parent company or oneof its subsidiaries, the value at which it was transferred tothe establishment was to be reported.

If an establishment changed ownership during the year,the cost of the fixed assets (building and equipment) wasto be reported.

QUANTITY OF ELECTRIC ENERGY CONSUMED FORHEAT AND POWER

Data on the quantity of purchased electric energy werecollected. In addition, information was collected on thequantity of electric energy generated by the establishmentand the quantity of electric energy sold or transferred toother establishments of the same company.

COST OF MATERIALS

This term refers to direct charges actually paid or payablefor items consumed or put into production during theyear, including freight charges and other direct chargesincurred by the establishment in acquiring these materials.It includes the cost of materials or fuel consumed,whether purchased by the individual establishment fromother companies, transferred to it from other establish-ments of the same company, or withdrawn from inventoryduring the year.

Included in this item are:

1. Cost of parts, components, containers, etc. In-cludes all raw materials, semifinished goods, parts,containers, scrap, and supplies put into production orused as operating supplies and for repair and mainte-nance during the year.

2. Cost of products bought and sold without fur-ther processing.

3. Cost of fuels consumed for heat and power. In-cludes the cost of fuel consumed, whether purchasedby the individual establishment from other companies,transferred to it from other establishments of thesame company, or withdrawn from inventory duringthe year.

4. Cost of purchased electricity. The cost of pur-chased electric energy represents the amount actuallyused during the year for heat and power. In addition,information was collected on the quantity of electricenergy generated by the establishment and the quan-tity of electric energy sold or transferred to otherplants of the same company.

Appendix A A–1Annual Survey of Manufactures

U.S. Census Bureau

5. Cost of contract work. This term applies to workdone by others on materials furnished by the manu-facturing establishment. The actual cost of the mate-rial was reported on the cost of materials, parts, andcontainers line of this item. The term ‘‘Contract Work’’refers to the fee a company pays to another companyto perform a service.

Duplication in Cost of Materials and Value ofShipments

The aggregate of the cost of materials and value of ship-ments figures for industry groups and for all manufactur-ing industries includes large amounts of duplication sincethe products of some industries are used as materials byothers. This duplication results, in part, from the additionof related industries representing successive stages in theproduction of a finished manufactured product. Examplesare the addition of flour mills to bakeries in the foodgroup and the addition of pulp mills to paper mills in thepaper and allied products group of industries. Estimates ofthe overall extent of this duplication indicate that thevalue of manufactured products exclusive of such duplica-tion (the value of finished manufactures) tends to approxi-mate two-thirds of the total value of products reported inthe survey.

Duplication of products within individual industries is sig-nificant within a number of industry groups; e.g., machin-ery and transportation industries. These industries fre-quently include complete machinery and their parts. Inthis case, the parts made for original equipment are mate-rials consumed for assembly plants in the same industry.

Even when no significant amount of duplication isinvolved, value of shipments figures are deficient as mea-sures of the relative economic importance of individualmanufacturing industries or geographic areas because ofthe wide variation in ratio of materials, labor, and otherprocessing costs of value of shipments, both amongindustries and within the same industry.

Before 1962, cost of materials and value of shipmentswere not published for some industries which includedconsiderable duplication. Since then, these data have beenpublished for all industries at the U.S. level and beginningin 1964, for all geographic levels.

VALUE OF SHIPMENTS

This item covers the received or receivable net selling val-ues, f.o.b. plant (exclusive of freight and taxes), of allproducts shipped, both primary and secondary, as well asall miscellaneous receipts, such as receipts for contractwork performed for others, installation and repair, sales ofscrap, and sales of products bought and sold without fur-ther processing. Included are all items made by or for theestablishments from materials owned by it, whether sold,transferred to other plants of the same company, or

shipped on consignment. The net selling value of productsmade in one plant on a contract basis from materialsowned by another was reported by the plant providing thematerials.

In the case of multiestablishment companies, the manu-facturer was requested to report the value of productstransferred to other establishments of the same companyat full economic or commercial value, including not onlythe direct cost of production but also a reasonable propor-tion of ‘‘all other costs’’ (including company overhead) andprofit.

Duplication in Cost of Materials and Value ofShipments

The aggregate of the cost of materials and value of ship-ments figures for industry groups and for all manufactur-ing industries includes large amounts of duplication sincethe products of some industries are used as materials byothers. This duplication results, in part, from the additionof related industries representing successive stages in theproduction of a finished manufactured product. Examplesare the addition of flour mills industry products to baker-ies in the food group and the addition of pulp mills indus-try products to paper mills in the paper and allied prod-ucts group of industries. Estimates of the overall extent ofthis duplication indicate that the value of manufacturedproducts exclusive of such duplication (the value of fin-ished manufactures) tends to approximate two-thirds ofthe total value of products reported in the annual survey.

Duplication of products within individual industries is sig-nificant within a number of industry groups; e.g., machin-ery and transportation industries. These industries fre-quently include complete machinery and their parts. Inthis case, the parts made for original equipment are mate-rials consumed for assembly plants in the same industry.

Even when no significant amount of duplication isinvolved, value of shipments figures are deficient as mea-sures of the relative economic importance of individualmanufacturing industries or geographic areas because ofthe wide variation in ratio of materials, labor, and otherprocessing costs to value of shipments, both amongindustries and within the same industry.

Before 1962, cost of materials and value of shipmentswere not published for some industries which includedconsiderable duplication. Since then, these data have beenpublished for all industries at the U.S. level and beginningin 1964, for all geographic levels.

PAYROLL

This item includes the gross earnings of all employees onthe payrolls of operating manufacturing establishmentspaid in the calendar year. Respondents were told theycould follow the definition of payrolls used for calculating

A–2 Appendix A Annual Survey of Manufactures

U.S. Census Bureau

the federal withholding tax. It includes all forms of com-pensation, such as salaries, wages, commissions, dis-missal pay, bonuses, vacation and sick leave pay, andcompensation in kind, prior to such deductions as employ-ees’ social security contributions, withholding taxes,group insurance, union dues, and savings bonds. The totalincludes salaries of officers of corporations; it excludespayments to proprietors or partners of unincorporatedconcerns. Also excluded are payments to members ofarmed forces and pensioners carried on the active payrollsof manufacturing establishments.

The census definition of payrolls is identical to that recom-mended to all federal statistical agencies by the Office ofManagement and Budget. It should be noted that this defi-nition does not include employers’ social security contri-butions or other nonpayroll labor costs, such as employ-ees’ pension plans, group insurance premiums, andworkers’ compensation.

PRODUCTION-WORKER HOURS

This item covers hours worked or paid for at the manufac-turing establishment, including actual overtime hours (notstraight-time equivalent hours). It excludes hours paid forvacations, holidays, or sick leave.

VALUE ADDED

This measure of manufacturing activity is derived by sub-tracting the cost of materials, supplies, containers, fuel,purchased electricity, and contract work from the value ofshipments (products manufactured plus receipts for ser-vices rendered). The result of this calculation is adjustedby the addition of value added by merchandising opera-tions (i.e., the difference between the sales value and thecost of merchandise sold without further manufacture,processing, or assembly) plus the net change in finishedgoods and work-in-process between the beginning- andend-of-year inventories.

For those industries where value of production is collectedinstead of value of shipments, value added is adjustedonly for the change in work-in-process inventoriesbetween the beginning and end of year. For those indus-tries where value of work done is collected, the valueadded does not include an adjustment for the change infinished goods or work-in-process inventories.

‘‘Value added’’ avoids the duplication in the figure forvalue of shipments that results from the use of productsof some establishments as materials by others. Valueadded is considered to be the best value measure avail-able for comparing the relative economic importance ofmanufacturing among industries and geographic areas.

EMPLOYEES

This item includes all full-time and part-time employees onthe payrolls of operating manufacturing establishmentsduring any part of the pay period which included the 12th

of the months specified on the report form. Included areall persons on paid sick leave, paid holidays, and paidvacations during these pay periods. Officers of corpora-tions are included as employees; proprietors and partnersof unincorporated firms are excluded. The ‘‘all employees’’number is the average number of production workers plusthe number of other employees in mid-March. The numberof production workers is the average for the payroll peri-ods including the 12th of March, May, August, and Novem-ber.

Production Workers

This item includes workers (up through the line-supervisorlevel) engaged in fabricating, processing, assembling,inspecting, receiving, storing, handling, packing, ware-housing, shipping (but not delivering), maintenance,repair, janitorial and guard services, product development,auxiliary production for plant’s own use (e.g., powerplant), recordkeeping, and other services closely associ-ated with these production operations at the establish-ment covered by the report. Employees above theworking-supervisor level are excluded from this item.

All Other Employees

This item covers nonproduction employees of the manu-facturing establishment, including those engaged in fac-tory supervision above the line-supervisor level. Itincludes sales (including driver-salespersons), sales deliv-ery (highway truck drivers and their helpers), advertising,credit, collection, installation and servicing of own prod-ucts, clerical and routine office functions, executive, pur-chasing, financing, legal, personnel (including cafeteria,medical, etc.), professional, and technical employees. Alsoincluded are employees on the payroll of the manufactur-ing establishment engaged in the construction of majoradditions or alterations utilized as a separate work force.

FRINGE BENEFITS

Fringe benefits consists of legally required expendituresand payments for voluntary programs. The legallyrequired portion consists primarily of federal old age andsurvivors’ insurance, unemployment compensation, andworkers’ compensation. Payments for voluntary programsinclude all programs not specifically required by legisla-tion whether they were employer initiated or the result ofcollective bargaining. They include the employer portionof such plans as insurance premiums, premiums forsupplemental accident and sickness insurance, pensionplans, supplemental unemployment compensation, wel-fare plans, stock purchase plans on which the employerpayment is not subject to withholding tax, and deferredprofit-sharing plans. They exclude such items as company-operated cafeterias, in-plant medical services, free parkinglots, discounts on employee purchases, and uniforms andwork clothing for employees.

Appendix A A–3Annual Survey of Manufactures

U.S. Census Bureau

TOTAL COMPENSATION

This term refers to the annual payroll or gross earningspaid in each calendar year to employees at the establish-ment. It includes all forms of compensation, such as sala-ries, wages, commissions, dismissal pay, bonuses, vaca-tion and sick leave pay, and compensation in kind, prior tosuch deductions as employees’ social security contribu-tions, withholding taxes, group insurance, union dues,and savings bonds. The total includes salaries of officers

of corporations; it excludes payments to proprietors orpartners of unincorporated concerns. Also excluded arepayments to members of armed forces and pensioners car-ried on the active payrolls of manufacturing establish-ments.

It should be noted that this definition does not includeemployers’ social security contributions or other nonpay-roll labor costs, such as employees’ pension plans, groupinsurance premiums, and workers’ compensation.

A–4 Appendix A Annual Survey of Manufactures

U.S. Census Bureau

Appendix B.Coverage and Methodology

DESCRIPTION OF THE ASM SURVEY SAMPLE

The ASM sample is selected at 5-year intervals. A newsample is introduced for the second survey year subse-quent to each census. Since 1999 was the second surveyyear following the 1997 census, a new sample wasselected based on the 1997 census.

In 1997, both the Census and ASM converted from the SICsystem to NAICS for purposes of assigning industrial clas-sifications to establishments. For the most part, the over-all scope of the manufacturing sector is similar under bothclassification systems; however, there are selected activi-ties that are classified as manufacturing under SIC that areconsidered as nonmanufacturing under NAICS. Thereverse is also true.

In 1997, there were approximately 366,000 individualmanufacturing establishments. For sample efficiency andcost considerations, the 1997 manufacturing populationwas partitioned into two components: a mail stratum anda nonmail stratum.

Mail Stratum

The mail stratum of the survey is comprised of largersingle-location manufacturing companies and all manufac-turing establishments of multiunit companies (companiesthat operate at more than one physical location). Approxi-mately 200,000 of the 366,000 establishments in the1997 census were assigned to the mail stratum.

On an annual basis, the mail stratum is supplementedwith larger, newly active single-location companies identi-fied from a list provided by the Internal Revenue Service(IRS) and new manufacturing locations of multiunit compa-nies identified from the Census Bureau’s Company Organi-zation Survey (COS).

For the 1999 survey, a new sample of approximately52,000 individual establishments was selected from themail stratum assembled from the 1997 census. Supple-mental samples representing both 1998 and 1999 births(newly active establishments that were not included in the1997 census) were also selected. Establishments selectedfor the sample will be mailed an ASM survey questionnairefor each year through 2003.

The 1999-2003 ASM sample design is similar to the oneused since 1984. Several adjustments were made, how-ever, to the arbitrary certainty portion of the sample. Com-panies in the 1997 Census of Manufactures with manufac-turing shipments of at least $1 billion were defined as

company certainties. (This cutoff was raised for 1999 tocompensate for inflation.) For these large companies, eachmanufacturing establishment is included in the mailsample. For the 1999-2003 sample, there are approxi-mately 500 certainty companies collectively accountingfor over 14,000 establishments.

For the remaining portion of the mail component of thesurvey, the establishment was defined as the sample unit.All establishments with 500 employees or more weredefined as employment certainties. (This was raised from250 in the 1994 sample.) In addition, all establishmentsproducing electronic computers were defined as certain-ties. A final certainty category, small industry certainties,was added for the 1999 sample. This category assuredthat all establishments of small industries (20 establish-ments or less) were included in the sample. Across thesefour arbitrary certainty classes, there were approximately16,600 establishments included in the sample with cer-tainty. Collectively, these certainty establishmentsaccounted for approximately 62 percent of the total valueof shipments in the 1997 Census of Manufactures.

Smaller establishments in the remaining portion of themail stratum were sampled with probabilities rangingfrom .02 to 1.00. The initial probabilities of selectionassigned to these establishments were proportionate to ameasure-of-size determined for each establishment. Themeasure-of-size was a function of the establishment’s1997 industry classification and its 1997 product classdata. For each product class and industry, a desired reli-ability constraint was specified. The product class con-straints were based on product class size (1997 ship-ments) while the industry constraints were a function ofboth current size (1997 shipments) and change in sizesince 1992. Using a technique developed by Dr. James R.Chromy of the Research Triangle Institute, the initial estab-lishment probabilities were optimized such that theexpected sample satisfied all industry and product classreliability constraints while the sample size was mini-mized. This technique reduces the likelihood of selectingnonrepresentative samples for individual product classesor industries.

This method of assigning probabilities based on productclass shipments is motivated by our primary desire to pro-duce reliable estimates of both product class and industryshipments. The high correlation between shipments andemployment, value-added, and other general statistics

Appendix B B–1Annual Survey of Manufactures

U.S. Census Bureau

assures that these variables will also be well representedby the sample. The actual sample selection procedureuses an independent chance of selection method (Poissonsampling).

Nonmail Component

The initial nonmail component of the ASM traditionallyconsists solely of small, single-establishment companiesfrom the census of manufactures. For 1999, the initialnonmail component also included approximately 3,600single-establishment companies that were rotated fromthe mail stratum to the nonmail stratum in order to pre-vent their selection in consecutive sample panels. Thecombined initial nonmail component of the 1999 sampleuniverse contained approximately 166,000 single-establishment companies from the 1997 Census of Manu-factures. The nonmail stratum is also supplemented annu-ally using the list of newly active single-locationcompanies provided by the IRS and payroll cutoffs. Com-panies with payroll below the payroll cutoff are added tothe nonmail stratum.

For this portion of the population, sampling is not used.The data for this group are estimated based on selectedinformation obtained annually from the administrativerecords of the IRS and Social Security Administration (SSA).This administrative information, which includes payroll,total employment, industry classification, and physicallocation, is obtained under conditions that safeguard theconfidentiality of both tax and census records.

DESCRIPTION OF THE ASM ESTIMATINGPROCEDURE

Most of the ASM estimates derived for the mail stratumare computed using a difference estimator. At the estab-lishment level, there is a strong correlation between thecurrent-year data values and the corresponding 1997(base) data values. Therefore, within the mailed stratum,for each item at each level of aggregation, an estimate ofthe ‘‘difference’’ between the current year and the baseyear is computed from sample cases and added to the cor-responding base-year values. For the 1999-2002 ASM esti-mates, the 1997 Census of Manufactures values serve asthe base year. For the 2004 ASM, the base will be updatedto the 2002 Economic Census - Manufacturing.

Because of the positive year-to-year correlation, estimatesderived using this methodology are generally more reli-able than comparable estimates developed from the cur-rent sample data alone. Estimates for the capital expendi-tures variables are not generated using the differenceestimator because the year-to-year correlations are consid-erably weaker. The standard linear estimator is used forthese variables.

For the nonmail stratum, estimates for payroll andemployment are directly tabulated from theadministrative-record data provided by IRS and SSA. Esti-mates of data other than payroll and employment are

developed from industry averages. Although the nonmailstratum contains approximately 202,000 individual estab-lishments in 1999, it accounts for only 2 percent of theestimate for total value of shipments at the total manufac-turing level.

Corresponding estimates for the mail and nonmail compo-nents are combined to produce the estimates included inthis publication.

QUALIFICATIONS OF THE ASM DATA

The estimates developed from the sample are apt to differsomewhat from the results of a survey covering all compa-nies in the sample lists but otherwise conducted underessentially the same conditions as the actual sample sur-vey. The estimates of the magnitude of the samplingerrors (the difference between the estimates obtained andthe results theoretically obtained from a comparable,complete-coverage survey) are provided by the standarderrors of estimates.

The particular sample selected for the ASM is one of manysimilar probability samples that, by chance, might havebeen selected under the same specifications. Each of thepossible samples would yield somewhat different sets ofresults, and the standard errors are measures of the varia-tion of all the possible sample estimates around the theo-retically comparable, complete-coverage values.

Estimates of the standard errors have been computedfrom the sample data for selected ASM statistics in thisreport. They are represented in the form of relative stan-dard errors (the standard errors divided by the estimatedvalues to which they refer).

In conjunction with its associated estimate, the relativestandard error may be used to define confidence intervals(ranges that would include the comparable, complete-coverage value for specified percentages of all the pos-sible samples).

The complete-coverage value would be included in therange:

• From one standard error below to one standard errorabove the derived estimate for about two-thirds of allpossible samples.

• From two standard errors below to two standard errorsabove the derived estimate for about 19 out of 20 of allpossible samples.

• From three standard errors below to three standarderrors above the derived estimate for nearly all samples.

An inference that the comparable, complete-survey resultwould be within the indicated ranges would be correct inapproximately the relative frequencies shown. Those pro-portions, therefore, may be interpreted as defining the

B–2 Appendix B Annual Survey of Manufactures

U.S. Census Bureau

confidence that the estimates from a particular samplewould differ from complete-coverage results by as muchas one, two, or three standard errors, respectively.

For example, suppose an estimated total is shown at50,000 with an associated relative standard error of 2 per-cent, that is, a standard error of 1,000 (2 percent of50,000). There is approximately 67 percent confidencethat the interval 49,000 to 51,000 includes the complete-coverage total, about 95 percent confidence that the inter-val 48,000 to 52,000 includes the complete-coveragetotal, and almost certain confidence that the interval47,000 to 53,000 includes the complete-coverage total.

In addition to the sample errors, the estimates are subjectto various response and operational errors: errors of col-lection, reporting, coding, transcription, imputation fornonresponse, etc. These operational errors also wouldoccur if a complete canvass were to be conducted underthe same conditions as the survey. Explicit measures oftheir effects generally are not available. However, it is

believed that most of the important operational errorswere detected and corrected during the Census Bureau’sreview of the data for reasonableness and consistency.The small operational errors usually remain. To someextent, they are compensating in the aggregated totalsshown. When important operational errors were detectedtoo late to correct the estimates, the data were suppressedor were specifically qualified in the tables.

As derived, the estimated standard errors included part ofthe effect of the operational errors. The total errors, whichdepend upon the joint effect of the sampling and opera-tional errors, are usually of the order of size indicated bythe standard error, or moderately higher. However, for par-ticular estimates, the total error may considerably exceedthe standard errors shown. Any figures shown in thetables in this publication having an associated standarderror exceeding 15 percent may be combined with higherlevel totals, creating a broader aggregate, which then maybe of acceptable reliability.

Appendix B B–3Annual Survey of Manufactures

U.S. Census Bureau