U.S. Department of Commerce Economics and Statistics Administration U.S. CENSUS BUREAU Annual Benchmark Report for Retail Trade Issued August 1999 BR/98-A U S C E N S U S B U R E A U Helping You Make Informed Decisions Current Business Reports January 1989 to December 1998 A Detailed Summary of Retail Sales, Purchases, Accounts Receivable, and Inventories
59
Embed
Annual Benchmark Report for Retail Trade: January 1989 to … · 2010. 8. 31. · Annual Benchmark Report for Retail Trade: January 1989 to December 1998 This publication contains
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
U.S. Department of CommerceEconomics and Statistics AdministrationU.S. CENSUS BUREAU
Annual Benchmark Reportfor Retail Trade
Issued August 1999
BR/98-A
U S C E N S U S B U R E A UHelping You Make Informed Decisions
Current Business Reports
January1989to December1998
A Detailed Summary of Retail Sales,Purchases, Accounts Receivable, and Inventories
Anne Russell, Assistant Division Chieffor Service Sector Statistics Division’sRetail and Wholesale Indicator programs,provided general direction in preparingthis report. Nancy Piesto, Chief, MonthlyRetail Surveys Branch, supervised the plan-ning, implementation, and compilation ofthe retail sales estimates with assistancefrom Vicki Garrett and Judy O’Neil.Patricia Borgstede, Gary Buffington,John Bywaters, Kelven Jones, SirettaKelly, Justin McLaughlin, ChristopherPece, Anthony Rivetti, Amelia Sharpe,Jean Swann, Sandra Wise, RussellYoung, Gwenn Byrne, Eleanor Daniels,and Lisa Kemp performed analyses andprocessing.
Ronald Piencykoski, Chief, CurrentRetail and Wholesale Indicators Branch,supervised the compilation of annualsales, inventories, purchases, andaccounts receivables with assistance fromDorothy Engleking, Scott Scheleur,Joan Brown, Judy Ross Davis, JamesErbacher, Louis Garner, Cheryl Geter,Eldridgina Houston, James Millet,Paul O’Donnell, Kathleen Denton, andMaggie Anderson.
Beverly Eng, Chief, Current Surveys Pro-cessing Branch, coordinated the collectionand processing of data through theNational Processing Center and the Eco-nomic Statistical Methods and Program-ming Division with assistance from JanisByrd, Will Caldwell, Debbie Dizebba,John Guest, Jennifer Juzwiak, Earnes-tine Kornegay, Lisa Endy-Mitchell,Elizabeth Murawski, Evelyn Shaw, andChristina Gatewood.
Ruth Detlefsen, Assistant Division Chieffor Service Sector Statistics Division’sResearch and Methodology, gave generaldirection to the statistical methodology.Jock Black, Chief, Program Research andDevelopment Branch, supervised thesample design, including estimation andvariance methodology with assistancefrom M. Cristina Cruz and WilliamDavie Jr.. Carol King, Chief, StatisticalMethods Branch, supervised the samplecontrol, imputation, and quality control
procedures with assistance from JamesBurton, James W. Hunt, JenniferJohnson, and Linda Moore. MichaelShimberg assisted in the seasonal adjust-ment and benchmarking of the estimatesin this publication.
Charles P. Pautler, Jr., Chief of EconomicStatistical Methods and Programming Divi-sion and Sarah W. Baumgardner, Assis-tant Division Chief, Current Indicators Pro-gram directed the development of thecomputer programs. Ronald M. Catzva,Chief, Service Sector Statistics DivisionIndicators Branch, directed the implemen-tation of monthly programs with assis-tance from Phillip Atwater, ChristineFord, Denise Gillis, Robert Janshego,Michael Lavarn, Steven Pirrone, andPatricia Young. Barry F. Sessamen,Chief, Service Sector Statistics AnnualBranch, directed the implementation ofannual programs with assistance fromKimberly Bankard, Jeffrey Burton,Tatiana Gallagher, and GarySchlegelmilch.
The staff of the Administrative and Cus-tomer Services Division, Walter Odom,Chief, performed planning, design, compo-sition, editorial review, and printing plan-ning and procurement for publications,Internet products, and report forms. HelenCurtis provided publication coordinationand editing.
Special acknowledgment is due for thecontributions made by Michael Lutz ofthe Economic Planning and CoordinationDivision, Larry Impett of the Service Sec-tor Statistics Division, and Carlene Bot-torff, Chief, Economic Projects Branch andall employees for the coordination, collec-tion, and processing of data in Jefferson-ville, Indiana’s National Processing Center.Special thanks are also given to the manybusinesses who provided data to the sur-vey. The survey would not have been pos-sible without their cooperation.
Direct all inquiries concerning this reportto Nancy Piesto, Service Sector StatisticsDivision, U.S. Census Bureau, Washington,DC 20233, telephone 301-457-2706.
ACKNOWLEDGMENTS
U.S. Department of CommerceWilliam M. Daley,
Secretary
Robert L. Mallett,Deputy Secretary
Economics and Statistics AdministrationRobert J. Shapiro,
Under Secretary for Economic Affairs
U.S. CENSUS BUREAUKenneth Prewitt,
Director
Annual Benchmark Reportfor Retail Trade
Issued August 1999
BR/98-A
Current Business Reports
January1989to December1998
A Detailed Summary of Retail Sales,Purchases, Accounts Receivable, and Inventories
Paula J. Schneider,Principal Associate Directorfor Programs
Frederick T. Knickerbocker,Associate Directorfor Economic Programs
Thomas L. Mesenbourg,Assistant Directorfor Economic Programs
Carole A. Ambler,Chief, Service SectorStatistics Division
SUGGESTED CITATION
U.S. Census Bureau, Current BusinessReports, Series BR/98-A, Annual
Benchmark Report for Retail Trade:January 1989 to December 1998,
Washington, DC 1999
For sale by Customer Services, U.S. Census Bureau, Washington, DC 20233.
ECONOMICS
AND STATISTICS
ADMINISTRATION
Economicsand StatisticsAdministration
Robert J. Shapiro,Under Secretaryfor Economic Affairs
TABLES1. Estimated Monthly Retail Sales by Kinds of Business, Unadjusted
and Adjusted: January 1989 to December 1998 . . . . . . . . . . . 22. Estimated End-of-Month Retail Inventories and Inventories/Sales
Ratios by Kinds of Business: January 1989 to December 1998 . . . 223. Estimated Purchases by Kinds of Business: 1989 to 1997 . . . . . . . 324. Estimated Gross Margin by Kinds of Business: 1989 to 1997 . . . . . 335. Estimated Gross Margin as Percent of Sales by Kinds of Business:
1989 to 1997 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346. Estimated Accounts Receivable of Retail Stores by Type of Account
and Kinds of Business: End of Year 1997 and 1996 . . . . . . . . . 357. Estimated Per Capita Sales by Selected Kinds of Business: 1989 to
Annual Benchmark Report for Retail Trade: January 1989 toDecember 1998
This publication contains the following retail trade esti-mates:
• Annual and monthly sales by retail stores at the nationallevel.
• End-of-year and end-of-month inventory values.
• Inventories/Sales ratios.
• Annual purchases and gross margin levels.
• Gross margin, as a percent of sales.
• Accounts receivables owed to retail stores.
• Per capita sales.
SUMMARY OF CHANGES
This publication presents estimates that reflect a bench-marking process that revised previously published unad-justed monthly sales and end-of-month inventory esti-mates for the period January 1993 through December1998. Estimates of annual sales, purchases, and end-of-year inventory values were revised for the years 1993through 1997. The revised unadjusted estimates reflect:
• Results from the 1992 Census of Retail Trade.
• Preliminary results from the 1997 Census of RetailTrade.
• Subsequent new annual estimates developed for theyears 1993 through 1996.
New factors were developed and used to adjust salesand inventories for the period January 1990 throughDecember 1998. The factors were used to adjust sales forseasonal, trading-day, and holiday variations and to adjustinventories for seasonal variations. These factors weredeveloped by using unadjusted estimates as input to theseasonal adjustment program. Adjusted estimates werederived beginning January 1990 instead of January 1993because the revised 1993 unadjusted estimates can affectthe computation of seasonal factors as far back as 1990.
The revision procedures were applied to specificdetailed sales and inventory series. All other sales andinventory series were obtained by summing appropriatedetail level estimates, thereby assuring additivity.
REVISIONS TO PREVIOUSLY PUBLISHED DATA
Annual Estimates
Annual estimates of sales, inventories, and purchasesderived for the years 1993 through 1997 were adjusted to1992 and 1997 Census of Retail Trade estimates1. The1993 through 1997 annual estimates of sales, end-of-yearinventories, and purchases were input to the benchmark-ing program and revised to—
• Equate the 1997 estimates to preliminary resultsderived from the 1997 Census of Retail Trade2.
• Link the series to the 1992 Census of Retail Trade level.
• Maintain, as best as possible, the year-to-year trendscomputed from estimates derived from our annual sur-vey for the years 1993 through 1997.
Monthly Sales
The final monthly sales sample estimates for December1992 through December 1998 were used as input to thebenchmarking program. The estimates were revised to—
• Constrain the sum of the 12 monthly estimates to equalthe revised annual estimates derived for the years 1993through 1997.
• Maintain, as best as possible, the month-to-month per-cent changes that were previously published.
The previously published December 1992 estimate wasalso used to link the revised estimates to the previouslypublished estimates. Also, prior to implementing thebenchmarking operation, corrections were applied to theestimates obtained directly from the sample for selectedkinds of business.
A mathematical result of the revision methodology isthat all sample estimates following the last benchmarkyear (1997) are revised by multiplying these estimates bythe ratio of the revised-to-sample estimate for December1997. Therefore, for specified detailed sales series, a ratioof the December 1997 revised estimate to the December1997 input estimate was computed. Sample estimates for
1The 1992 and 1997 sales were adjusted to include an estimate for nonemployerdirect sellers (SIC 5963). These estimates were obtained from the 1992 and 1997 AnnualRetail Trade Surveys.
2The inventory and purchase estimates for the census years 1992 and 1997 werecomputed by multiplying the inventories (or purchases) obtained from the 1992 and 1997annual surveys by the ratio of preliminary census sales to sales derived from the 1992 and1997 annual surveys. The procedure was implemented because neither inventory norpurchase data were collected in the census.
v
U.S. Census Bureau
months following December 1997 were multiplied bythese constant ratios (called carry-forward factors) toderive revised sales estimates. The carry-forward factorsremain constant until the next benchmarking.
End-of-Month Inventories
The final end-of-month inventory sample estimates forDecember 1992 through December 1998 were used asinput to the benchmarking program. The estimates wererevised to—
• Equate the December or end-of-year inventory estimatesfor the years 1993 through 1997 to the revised end-of-year annual estimates.
• Maintain, as best as possible, the month-to-month per-cent changes that were previously published.
The previously published December 1992 estimate wasalso used to link the revised estimates to the previouslypublished estimates. Also, prior to the implementation ofthe benchmarking operation, corrections were applied toestimates obtained directly from the sample for selectedkinds of business.
For specified detailed inventory series, revised esti-mates subsequent to December 1997 were derived bymultiplying the sample estimates by the ratio of therevised December 1997 estimate to the December 1997sample estimate. This ratio is the carry-forward factor forinventory estimates and is constant for all series until thenext benchmarking.
REVISIONS OF SALES AND INVENTORIES : 1998
The following table shows a comparison of the revisedsales and inventories estimates to the previously pub-lished estimates for 1998:
1998 Retail Sales and Inventories Comparison of theRevised Estimates to the Previously Published Estimates(Millions of dollars)
1End of December. Not adjusted for seasonal variations.
Reasons for Monthly Revisions
There are several reasons for revisions. The main con-tributors to the revision from the previously publishedestimates are:
Changes in Classification. The sales and inventory esti-mates are benchmarked to preliminary results from the1997 Census of Retail Trade. Changes in classificationbetween the 1992 and 1997 censuses affect the levels ofthe monthly and annual estimates. The previously pub-lished estimates were benchmarked to the 1992 census,but the revised levels are benchmarked to the 1997 cen-sus. Classification differences may occur because 1)estab-lishments previously classified in other trade areaschanged to retail; 2) establishments previously classifiedas retail are now classified in another trade area; and 3)establishments changed classification within retail trade.
vi
U.S. Census Bureau
Timing. The respondents have more time to preparetheir annual reports than they do for their monthlyreports. The annual responses are requested at a timewhen many firms have already compiled audited book fig-ures for their own use. The timing of the annual survey issuch that it is possible to obtain independent verificationof the reported data from such sources as a company’sannual report. On the other hand, respondents to themonthly survey have just a few weeks to provide reportsof their sales and end-of-month inventories. Sometimesthese reports are based on incomplete or unauditedrecords or include estimates by the respondents to repre-sent their understanding of their business.
Sampling. The annual sample is larger than the samplesused to develop the estimates for any given month. Asdescribed in more detail in Appendix A of this report, thesamples used to produce the sales estimates each monthconsist of certainty companies and noncertainty EmployerIdentification Numbers (EINs).
• The companies selected as certainty companies arethose companies that had large sales in relation to othercompanies in their kind-of-business group.
• The initial selection of the noncertainty EINs was basedon stratified random sampling, using annual sales asthe measure of size for those EINs not belonging to acertainty company. The selected noncertainty EINs wereassigned to one of two groups. The monthly samplecanvasses one group of noncertainty EINs while bothgroups of noncertainty EINs are canvassed in the annualsample.
• The sample used for the end-of-month inventory esti-mates is a sub-sample of the monthly sales sample. Themonthly inventories sample is about one-third the sizeof the monthly sales sample. This, combined with thedifferences between the annual and monthly surveysmentioned above, often results in greater revisions toinventory estimates than sales estimates.
Response. The annual estimates are based on morereported data than are the monthly estimates. Theresponse to the Annual Retail Trade Survey is required bylaw. This requirement results in a dollar volume responserate above 90 percent. The response to the monthly sur-vey is not mandatory. The response rates for the monthlysurveys are usually around 75 percent for sales and 70percent for inventories. The sales and inventories for thenon-reported retailers are accounted for by an imputationprocess. This process assumes that the nonrespondingfirms have trends similar to the responding firms, in theirrespective kinds of business.
Adjustment Factors
In addition to the benchmarking operation, new factorswere developed to adjust the 1990 through 1998 esti-mates of monthly sales for seasonal variations as well as
trading-day and holiday differences. Unadjusted sales esti-mates for the months of January 1972 through May orJune 1999 were used as input. New factors were devel-oped to adjust the 1990 through 1998 end-of-monthinventory estimates, using the months of December 1980through May 1999 as input.
The seasonal adjustment program which developed thefactors in this publication was the X-11 ARIMA version,developed by Statistics Canada. Using forecasting, theadjustment factors were developed for the sales of depart-ment stores (SIC 5311), variety stores (SIC 5331), and mis-cellaneous general merchandise stores (SIC 5399) by theversion using forecasting. All other factors were calculatedwithout forecasting.
DEFINITION OF TERMS
Retail trade, as defined by Major Groups 52 through 59of the 1987 Standard Industrial Classification Manualincludes establishments that are engaged in the sale ofmerchandise for personal or household consumption andin the rendering of services that are incidental to the saleof the merchandise. Additional characteristics of retailestablishments are:
• They usually operate at fixed places of business.
• They are engaged in activities to attract the generalpublic to buy.
• They buy or receive and sell merchandise.
• They may process their products, but such processing isincidental or subordinate to selling.
• They are considered as retail in the trade.
Sales
Sales include merchandise sold (for cash or credit atretail or wholesale) by establishments primarily engagedin retail trade. Services that are incidental to the sale ofmerchandise, and excise taxes that are paid by the manu-facturer or wholesaler and passed along to the retailer arealso included. Sales are net after deductions for refundsand allowances for merchandise returned by customers.Sales exclude sales taxes collected directly from custom-ers and paid directly to a local, state, or federal taxagency.
The estimates of sales measure the operations receiptsrendered by stores that primarily sell at retail. The salesestimates represent total sales and receipts of all estab-lishments primarily engaged in retail trade. They do notinclude sales at retail by manufacturers, wholesalers, serv-ice establishments, and others whose primary activity isother than retail trade. Because the retail establishment isthe basic unit of measure, the published estimates of salesby type of retail store are not intended to measure thetotal sales for a given commodity or merchandise line.
vii
U.S. Census Bureau
Inventories
Merchandise inventories are the value of stocks ofgoods held for sale through retail stores. The inventoriesestimates represent the value, at cost, of the merchandiseavailable for sale as of the last day of the report period.Methods of valuation may vary according to the account-ing practices of each firm. The estimates provided in thisreport are valued on a non-LIFO (last in, first out) basis.Note - LIFO is a method of valuing inventory where the lat-est items of merchandise added to the inventory are thefirst ones taken out. Non-LIFO would mean that anothermethod, such as FIFO (first in, first out), was used toestablish the value of the inventory available for sale. Mer-chandise inventories are shown for stores and warehousesservicing retail establishments. Included are only thosewarehouses that maintained supplies of merchandise pri-marily intended for distribution to retail stores within theorganization.
In the Annual Retail Trade Survey, most firms reportedthe value of their inventories as of the close of the calen-dar year. Some firms, using a fiscal year rather than a cal-endar year for accounting purposes, valued their inven-tory as of some date other than the last day of thecalendar year. About 26 percent of the retail inventoryestimate from the Annual Retail Trade Survey was basedon data reported on an end-of-fiscal year other thanDecember 31. In the annual survey, inventories reportedfor a date in a month other than December were adjustedto a December 31 equivalent, based on ratios developedfrom the monthly inventory sample.
Inventories/Sales Ratios
The inventories/sales ratios show the relationship ofthe end-of-month values of inventories to the monthlysales. These ratios can be looked at as indications of thenumber of months of inventory that are on hand in rela-tion to the sales for a month. For example, a ratio of 2.5would indicate that the retail stores have enough mer-chandise on hand to cover two and a half months of sales.
Purchases
Purchases represent the total cost of merchandise thatwas purchased for resale during the year, whether or notpayment for the merchandise was made during the year.Purchases data include cash and credit purchases made atcentral offices and company warehouses. The purchases,however, exclude deliveries from central offices or ware-houses to retail stores owned by other companies. Thosecompanies engaged in both manufacturing and retailingare asked to include purchases at the cost value of inter-company transfers from their plant or warehouses to theirretail stores. Also, these companies are asked to includethe cost of outside purchases.
Cost of Goods Available for Sale and Cost of GoodsSold
To calculate the cost of goods available for sale, thepurchases made during the year are added to the preced-ing year’s end-of-year inventories. To calculate the cost ofgoods sold, the end of the current year’s value of invento-ries are deducted from the cost of goods available for sale.
Gross Margin
The measure of gross margin represents total sales lesscost of goods sold.
Accounts Receivable Balances
Retail accounts receivable are amounts owed to retailstores by their customers for purchases made on credit.The estimate in this publication refer to receivables out-standing as of December 31, including receivables againstwhich the firm has borrowed. However, credit paper dis-counted or sold to others and amounts actually chargedoff as bad debts are excluded. Also excluded are amountscharged on credit cards issued by oil companies, banks,and other organizations that issue credit cards.
Charge accounts are credit accounts for which full pay-ment is scheduled to be made at the end of the customarybilling period. Installment accounts are classified as open-end or closed-end.
• Open-end installment accounts are primarily revolvingor optional accounts. A deferred payment privilege isextended through a line of credit, and the customer hasthe option of paying the balance in full or paying in twoor more installments. The payments are subject to aminimum required payment with a finance charge usu-ally assessed.
• Closed-end installment accounts are those generallyrequiring a new contract to cover each extension ofcredit. A pre-computed finance charge is assessed at thetime credit is extended. Specified fixed schedules ofinstallment payments are established with the number,the amount of payments, and the due dates specified inthe contract.
Leased Department
Leased departments (concessions) are broadly definedas operations of one company conducted within the estab-lishment of another company. Typical examples mayinclude jewelry counters or optical centers within depart-ment stores.
GAF
GAF represents sales at stores that sell merchandisenormally sold in department stores. GAF includes the fol-lowing kinds of retail businesses:
• General merchandise group stores (SIC 53).
viii
U.S. Census Bureau
• Apparel and accessory stores (SIC 56).
• Furniture group stores (SIC 57).
• Miscellaneous shopping goods stores (SIC 594).
Price Changes
The estimates presented in this publication are notadjusted for price changes.
ix
U.S. Census Bureau
Detailed Tables
1
U.S. Census Bureau
Table 1. Estimated Monthly Retail Sales by Kinds of Business, Unadjusted and Adjusted:January 1989 to December 1998
[Data in millions of dollars]
SIC code Kind of business1998
Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Total
1Includes data for leased departments operated within department stores. Data for this line not included in broader kind-of-business totals.2Includes establishments primarily selling merchandise through coin-operated vending machines, by house-to-house canvass, or from mail order.3GAF represents stores which specialize in department store types of merchandise (general merchandise, apparel, furniture, miscellaneous shopping goods stores).4Data adjusted for seasonal variations, holiday, and trading-day differences, but not for price changes.5For yearly totals, refer to unadjusted section.
Note: U.S. and group totals include kinds of business not shown.
21
U.S. Census Bureau
Table 2. Estimated End-of-Month Retail Inventories and Inventories/Sales Ratios by Kinds of Business:January 1989 to December 1998
[Data in millions of dollars]
SIC code Kind of business1998
Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.
1Inventory data adjusted for seasonal variations; sales data adjusted for seasonal variations, holiday, and trading-day differences, but not for price changes.
Note: U.S. and group totals include kinds of business not shown.
31
U.S. Census Bureau
Table 3. Estimated Purchases by Kinds of Business: 1989 to 1997
[Data in millions of dollars, not adjusted for price changes]
SIC code Kind of business 1989 1990 1991 1992 1993 1994 1995 1996 1997
1GAF represents stores which specialize in department store types of merchandise (general merchandise, apparel, furniture, and miscellaneous shopping goods stores).
Note: U.S. and group totals include kinds of business not shown.
32
U.S. Census Bureau
Table 4. Estimated Gross Margin by Kinds of Business: 1989 to 1997
[Data in millions of dollars, not adjusted for price changes]
SIC code Kind of business 1989 1990 1991 1992 1993 1994 1995 1996 1997
1GAF represents stores which specialize in department store types of merchandise (general merchandise, apparel, furniture, and miscellaneous shopping goods stores).
Note: U.S. and group totals include kinds of business not shown See introductory text for an explanation of gross margin. Because of rounding differences, the published gross margin may notequal the gross margin calculated from its published components.
33
U.S. Census Bureau
Table 5. Estimated Gross Margin as Percent of Sales by Kinds of Business: 1989 to 1997
SIC code Kind of business 1989 1990 1991 1992 1993 1994 1995 1996 1997
1GAF represents stores which specialize in department store types of merchandise (general merchandise, apparel, furniture, and miscellaneous shopping goods stores).
Note: U.S. and group totals include kinds of business not shown.
34
U.S. Census Bureau
Table 6. Estimated Accounts Receivable of Retail Stores by Type of Account and Kinds of Business: End ofYear 1997 and 1996
(S) Does not meet publication standards because of high sampling or nonsampling error.1Includes mail-order houses engaged in the sale of department store merchandise (SIC 5961 part).
Note: Measures of sampling variability are shown in table A-3.
Table 7. Estimated Per Capita Sales by Selected Kinds of Business: 1989 to 1998[Dollars]
SIC code Kind of business 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
Note: Civilian population estimates (in thousands) as of July 1: 1989 — 245,131; 1990 — 247,799; 1991 — 250,513; 1992 — 253,417; 1993 — 256,280; 1994 — 258,880; 1995 — 261,409; 1996— 263,893; 1997 — 266,384; 1998 — 269,078
U.S. Department of Commerce, U.S. Census Bureau, Population Division, Population Estimates Program.The data are subject to sampling variability. However, measures of sampling variability for per capita sales are not available.
35
U.S. Census Bureau
Appendix A.Explanatory Material
INTRODUCTION
The U.S. Census Bureau produces the Annual Bench-mark Report for Retail Trade to provide national estimatesof annual sales and end-of-year inventories of retail estab-lishments by kind of business. The U.S. Census Bureauuses the more accurate annual estimates to revise the pre-viously published estimates of monthly sales and end-of-month inventories. These revisions are also included inthis report.
The sales and inventory estimates in this report aredeveloped using data from a probability sample. Thesample is taken from a universe of employer firms of allsizes and kinds of business in retail trade throughout thenation. Knowing each unit’s probability of being includedin the sample allows us to evaluate the sampling variabil-ity of the estimates.
SAMPLE DESIGN
This section describes the design of the sample used toestimate sales (monthly and annual) and end-of-yearinventories. The sample used to estimate end-of-monthinventories is a subsample of this sample. A description ofthe methods used to design this subsample is included atthe end of this section.
Sampling Frame
A new sample was used to compute estimates begin-ning with the April 1997 data month. This sectiondescribes the design and selection procedures for the newsample. For a description of the prior sample see theAnnual Benchmark Report for Retail Trade for 1997 orprior years.
The same sampling frame was used for the MonthlyRetail Trade Survey (MRTS) and the Annual Retail TradeSurvey (ARTS). The frame has two types of sampling unitsrepresented — Employer Identification Numbers (EINs) andlarge, multiple-establishment firms. Both sampling unitsrepresent clusters of one or more establishments ownedor controlled by the same firm. All of the information usedto create these sampling units was extracted from estab-lishment records contained on the Census Bureau’s Stand-ard Statistical Establishment List (SSEL) as updated toDecember 31, 1994. The next few paragraphs give detailsabout the SSEL; the distinction between firms, EINs, andestablishments; and the construction of the samplingunits used in the sample selection. Though important,
they are not essential to understanding the basic sampledesign and readers may continue to the Stratification,Sampling Rates, and Allocation section.
The SSEL is a multi-relational database that contains arecord for each establishment with employees. The estab-lishment is the smallest entity represented on the SSEL. Anestablishment is a single physical location where businesstransactions take place and for which payroll and employ-ment records are kept. Groups of one or more establish-ments under common ownership or control are firms. Asingleunit firm owns or operates only one establishment,whereas a multiunit firm owns or operates two or moreestablishments. The treatment of establishments on theSSEL differs according to whether the establishment is partof a multiunit or singleunit firm. In particular, the structureof an establishment’s primary identifier on the SSEL differsdepending on whether it is owned by a singleunit or multi-unit firm.
A singleunit firm’s primary identifier is its EmployerIdentification Number. The Internal Revenue Service (IRS)issues the EIN and the firm uses it as an identifier toreport social security payments for its employees underthe Federal Insurance Contributions Act (FICA). The sameact requires all employer firms to use EINs. Because single-unit firms have only one establishment, there is a one-to-one relationship between the firm and the EIN. Thus thefirm, the EIN, and the establishment all reference the samephysical location and all three terms can be used inter-changeably and unambiguously when referring to singleestablishment firms.
For multiunit firms however, a different structure con-nects an employer firm with its establishments via the EIN.When reading the following, it may help to keep in mindthe analogy of a filing cabinet (multiunit firm) composedof many drawers (EINs) each containing several files(establishments).
Essentially a multiunit firm is associated with a clusterof one or more EINs and EINs are associated with two ormore establishments. Each firm is associated with at leastone EIN and only one firm can use a given EIN. However,one firm may have several EINs. Similarly, there is a one-to-many relationship between EINs and establishments.Each EIN can be associated with many establishments buteach establishment is associated with only one EIN.Because of the possibility of one-to-many relationships, we
A-1
U.S. Census Bureau
must distinguish between the firm, its EINs, and its estab-lishments. The firm that owns or controls a multiunitestablishment is identified on the SSEL by way of theestablishment’s primary identifier.
A multiunit establishment’s primary identifier consistsof a unique combination of an alpha number and a plantnumber. The alpha number identifies the firm; and theplant number identifies the establishment within the firm.All establishments owned or controlled by the same firmhave the same alpha number. Different firms have differ-ent alpha numbers and different establishments within thesame firm have different plant numbers. The U.S. CensusBureau assigns both the alpha number to the firm and theplant numbers to the establishments based on the resultsof the quinquennial economic censuses and the annualCompany Organization Surveys.
To create the sampling frame, we extracted the recordsfor all establishments that have a Standard Industrial Clas-sification (SIC) that falls in the Retail Trade area as definedin the 1987 Standard Industrial Classification Manual. Forthese establishments, we extracted sales (or receipts),payroll, employment, inventory, name and address infor-mation, as well as primary identifiers and, for multiunitestablishments, associated EINs. We summarize the eco-nomic data of multiunit establishments to an EIN level bytabulating the establishment data for all retail establish-ments associated with the same EIN. Similarly we summa-rize to a firm level by tabulating the establishment datafor all establishments associated with the same alphanumber. These are the sampling units created from multi-unit establishments. No aggregation is necessary to putsingleunit establishment information on an EIN basis or afirm basis. Thus, the sampling units created for singleunitfirms simultaneously represent establishment, EIN, andfirm information. In summary, the sampling frame is acomplex amalgam of establishments, EINs, and firms.
Stratification, Sampling Rates, and Allocation
The primary stratification of the frame is by kind-of-business group based on the detail required for this publi-cation. We further stratify the sampling units within kindof business groups (substratify) by a measure of sizerelated to their sales. To reduce the variance of the esti-mates, the largest sampling units are selected ‘‘with cer-tainty.’’ This means they are sure to be selected and willrepresent only themselves (i.e., have a selection probabil-ity of one and a sampling weight of one). Within eachkind-of-business stratum a substratum boundary (or cut-off) that divides the certainty units from the noncertaintyunits is determined. We based these cutoffs on a statisticalanalysis of data from the 1992 Census of Retail Trade.Accordingly, these values are on a 1992 sales basis. Wealso used this analysis to set sampling rates needed toachieve specified sampling variability objectives for salesestimates for different kind-of-business groups. The recip-rocal of the sampling rate is referred to as the sampling
weight. Note that we computed sampling rates using datafrom the 1992 Census, the actual sample size was deter-mined by applying these sampling rates to the samplingframe constructed from the 1994 SSEL. We then allocatedthe sample optimally based on the number of samplingunits and standard deviation of the units’ measures ofsize. The allocation is optimal in the sense that it achievesthe minimum variance for a fixed sample size.
Sample Selection
The first step in the sample selection identified cer-tainty firms. If a firm had total retail sales (for 1994adjusted to a 1992 basis) greater than the correspondingcertainty cutoff for its major kind of business, the firmwas selected into the sample with certainty. In this case,the firm is considered the sampling unit. This has impor-tant consequences when the firm adds establishments. If afirm was selected with certainty and had more than oneestablishment at the time of sampling, any new establish-ments that the firm acquires, even if under new or differ-ent EINs, are included in the sample with certainty. If asingleunit firm was selected with certainty all future estab-lishments associated with that firm’s original EIN areincluded with certainty; any new EINs that might later beassociated with that firm are subjected to sampling asbirths (see below).
All firms not selected with certainty were subjected tosampling on an EIN basis. If a firm had more than one EIN,each of its EINs was a separate sampling unit. To be eli-gible for the initial sampling, an EIN had to have nonzeropayroll in 1994. The EINs were stratified according to theirmajor kind of business and their estimated sales (on a1992 basis). Within each noncertainty stratum, a simplerandom sample of EINs was selected. We then assignedthe selected cases to one of two groups. One group is can-vassed for both the monthly and annual surveys, the otheronly for the annual survey. The sampling weights for theEINs selected in the sample for the monthly sales surveyvaried between 1 in 1 and 1 in 968.6. The maximumweight for the noncertainty sample canvassed for theannual survey was 464.3.
Sample Maintenance
Periodically, we update the sample to represent EINsissued since the last sample selection. For all EIN ‘‘births,’’a two-phase selection procedure is used. EIN births arenew EINs recently assigned by the IRS, on the latest avail-able IRS mailing list for FICA taxpayers, and assigned akind-of-business classification (if possible) by the SocialSecurity Administration (SSA). In the first phase, births arestratified by kind of business and size (expected employ-ment or quarterly payroll). A relatively large sample isdrawn and canvassed to obtain a more reliable measure ofsize, consisting of sales in two recent months, and a moredetailed kind-of-business code, if needed.
A-2
U.S. Census Bureau
Using this more reliable information, the selected birthsfrom the first phase are subjected to probability-proportional-to-size sampling with overall probabilitiesequivalent to those used in drawing the initial samplefrom the 1994 SSEL. Because of the lag in reporting birthsto the IRS, and the time needed to accomplish the two-phase birth-selection procedure, births are added to thesample about nine months or more after they begin opera-tion.
The processing of the EIN births differs between themonthly and annual surveys. For the monthly survey(MRTS), the EIN births selected in a quarter are added intothe survey during the quarter. Because births are not rep-resented in the monthly sample until they go through thetwo-phase selection procedure, an interim procedure isused to account for births during the period between theonset of activity and the time of birth selection. This con-sists of imputing data for all selected cases that go out ofbusiness but are still on the IRS mailing list.
For the annual survey (ARTS), the births that areselected in the quarterly birth-selection procedure inNovember of the annual survey year are included in themain mailing of the ARTS questionnaires in January of thefollowing year. To better represent all EIN births in theannual survey year, and specifically to account for the cov-erage lag in the birth-selection procedure, we add birthsto the annual sample that are selected in February, May,and August of the year following the survey year. We mailsurvey forms to these births in June and August to supple-ment the main annual survey mailing.
To be eligible for the sample canvass and tabulation ina given month, a retail EIN selected in the noncertaintysampling operations must meet both of the followingrequirements:
• It must be on the latest available IRS mailing list forFICA taxpayers from the previous quarter.
• It must have been selected from either the SSEL or thefile of employer births.
EINs selected into the sample with certainty are notdropped from canvass and tabulation if they are no longeron the IRS mailing list. Rather, the business that used theEIN is contacted, and if a successor EIN is found, it isadded to the survey. This is a more stringent quality con-trol used for these larger businesses.
Estimation Procedures for Monthly Sales
All sampling units selected with certainty (weight equalto 1.0) are canvassed each month to obtain sales andother information for the month just ending. As noted pre-viously, two samples of noncertainty (weight greater than1.0) EINs were selected. Each month, all certainty sam-pling units and half of all selected noncertainty samplingunits are asked to report their sales data for the monthjust ending. (The remaining half of the noncertainty units
are canvassed only for the annual survey.) The estimatesare computed as the sum of weighted data (reported andimputed), where the weight for a given sampling unit isthe reciprocal of its probability of selection.
Estimation Procedures for Annual Data
The annual sales, purchases, end-of-year inventories,accounts receivable, and gross margin estimates pub-lished in this report are based on the ARTS. The estimatesare computed as the sum of weighted data (reported andimputed) for all selected sampling units that meet thetabulation criteria given in the Sample Maintenance sec-tion. The weight for a given sampling unit is the reciprocalof its probability of selection. For noncertainty units,annual weights are equal to half of the monthly weightbecause units from both noncertainty panels are used toproduce annual estimates whereas only one panel is usedto produce monthly estimates. The use of a larger sampleresults in annual estimates having less sampling variabil-ity than monthly estimates.
The estimates of total inventory published in this reportare on a non-LIFO basis. For those firms that value inven-tory on a LIFO (last-in, first-out) basis, the LIFO reserveamount is added to the LIFO value to get inventory on anon-LIFO basis. In the ARTS, sampling units that used aLIFO cost basis for all or part of their inventory were askedto report LIFO reserves.
The estimates of gross margin represent sales minusthe cost of goods sold (see text, Definition of Terms) andare a function of the sums of weighted sales, inventories,and purchases data reported by the sampling units in theARTS. For 1997 we adjusted the inventories and purchasesby multiplying the 1997 ARTS estimates by the ratio of1997 preliminary census sales to 1997 ARTS sales. There-fore, 1997 purchases and inventories are comparable tothe 1997 preliminary census sales shown in this report.
SUBSAMPLE TO ESTIMATE END-OF-MONTHINVENTORIES
As stated above, the sample used to estimate end-of-month inventories is a subset of the units used to esti-mate monthly and annual sales and end-of-year inventories.This section highlights differences between the design ofthe subsample and the complete sample.
Sampling Frame
The inventory sample is a subsample of the salessample. Thus, the inventory sampling frame is the salessample and has the same types of sampling units as thesales frame - companies and EINs. Descriptions of thesesampling units are given above.
Stratification, Sampling Rates, and Allocation
Sampling units are stratified by their major kind of busi-ness. The stratification used for the monthly inventorysample is based on groupings of 3-digit and 4-digit Stand-ard Industrial Classification codes. Because the estimates
A-3
U.S. Census Bureau
of end-of-month inventories are produced for somewhatbroader kind-of-business groupings, the stratification forthe inventory sample is coarser than the stratificationused for the sales sample. Sampling units are assigned tosubstrata based on a measure of size related to theirsales. Sampling rates for the inventory subsample arecomputed using the same methods as described previ-ously. The reciprocal of the sampling rate is referred to asthe sampling weight. Units tabulated in the monthly inven-tory survey have both an inventory weight and a salesweight. Because the inventory sample is a subsample ofthe sales sample, the inventory weight is greater than orequal to the sales weight.
Sample Selection
The selection of the inventory sample is carried outindependently within each sales-size substratum con-tained in a kind-of-business stratum. The actual selectionprocedure follows a systematic, probability-proportional-to-size scheme and will not be described here. Becausethe inventory sample is a subsample of the sales sample,it is possible that some units already have been selectedwith a sales weight that is greater than the desired inven-tory weight. These units are assigned to a separate panelwithout being subjected to a second sampling procedure.For these units, the inventory weight is set equal to thesales weight. Thus, in addition to the panel of certaintyunits and two panels of noncertainty units, there isanother panel of noncertainty units that was not subjectedto the subsampling. The sampling rates for the EINsselected in the sample for the monthly inventory surveyvaried between 1 in 1 and 1 in 2631.6.
Sample Maintenance
The inventory sample is updated in the same mannerand at the same time as the Monthly Retail Trade Surveysample. A subsample of the births selected for the salessample is selected for the inventory sample. The samplingrates of the initial inventory sampling are maintained.
Monthly Estimation Procedures
The procedures described above in the EstimationProcedures for Monthly Sales are followed except thesales weight is replaced by the inventory weight for allinventory computations.
RELIABILITY OF THE ESTIMATES
An estimate based on a sample survey potentially con-tains two types of errors - sampling and nonsampling.Sampling error occurs because characteristics differamong sampling units and because only a subset of theentire population is measured in a sample survey. Non-sampling error encompasses all other factors that contrib-ute to the total error of a sample survey estimate. Theaccuracy of a survey result may be affected by these twotypes of errors.
Sampling and nonsampling errors are often measuredby the quantities, bias and variance. The bias of an esti-mator of an unknown population value is the difference,averaged over all possible samples of the same size anddesign, between the estimator and the unknown popula-tion value. Any systematic error or inaccuracy that affectsall samples in a similar way will cause the resulting esti-mates to be biased. Variance is the squared difference,averaged over all possible samples of the same size anddesign, between an estimator and its average value.
Descriptions of sampling variability and nonsamplingerror for the Monthly Retail Trade Survey (MRTS) are pro-vided in the following sections.
Measures of Sampling Variability
Because the estimates are based on a sample, exactagreement with the results that would be obtained from acomplete enumeration of retail firms on the samplingframe is not expected. However, because each firm repre-sented on the sampling frame has a known probability ofbeing selected into the sample, it is possible to estimatethe sampling variability of the survey estimates.
The particular sample used in this survey is one of alarge number of samples of the same size that could havebeen selected using the same design. If all possiblesamples had been surveyed, under the same conditions,an estimate of an unknown population value could havebeen obtained from each sample. These samples give riseto a distribution of estimates for the unknown populationvalue. A statistical measure of the variability among theseestimates is the standard error, which can be approxi-mated from any one sample. The standard error isdefined as the square root of the variance. The coefficientof variation (or relative standard error) of an estimate isthe standard error of the estimate divided by the estimate.Note that measures of sampling variability, such as thestandard error or coefficient of variation, are estimatedfrom the sample and are also subject to sampling variabil-ity. (Technically, we should refer to the estimated standarderror or the estimated coefficient of variation of an estima-tor. However, for the sake of brevity, we have omitted thisdetail.) It is important to note that the standard error andcoefficient of variation only measure sampling variability.They do not capture any systematic biases in the esti-mates. The table Estimated Coefficients of Variation andStandard Error for Retail Sales by Kind of Business givesthe estimates of coefficients of variation in percent forrecent dollar-volume sales estimates and trends. The tableEstimated Coefficients of Variation and Standard Error forRetail Inventories by Kind of Business gives the estimatesof coefficients of variation in percent for recent monthlydollar-volume inventory estimates and trends.
The estimate from a particular sample and the approxi-mate standard error associated with the estimate can beused to construct a confidence interval. A confidence inter-val is a range about a given estimator that has a specified
A-4
U.S. Census Bureau
probability of containing the estimator’s corresponding,unknown population value. If, for each possible sample,an estimate of an unknown population value and itsapproximate standard error were obtained, then:
• For approximately 90 percent of the possible samples,the interval from 1.65 standard errors below to 1.65standard errors above the estimate would include theunknown population value.
• For approximately 95 percent of the possible samples,the interval from two standard errors below to twostandard errors above the estimate would include theunknown population value.
Nonsampling Errors
Nonsampling error encompasses all other factors thatcontribute to the total error of a sample survey estimateand may also occur in censuses. It is often helpful to thinkof nonsampling error as arising from deficiencies or mis-takes in the survey process. In the MRTS nonsamplingerror can be attributed to many sources: (1) inability toobtain information about all firms in the sample, (2)
response errors, (3) differences in the interpretation of thequestions, (4) mistakes in coding or keying the dataobtained, and (5) other errors of collection, response, cov-erage, and processing. Although no direct measurement ofthe potential biases due to nonsampling error has beenobtained, precautionary steps were taken in all phases ofthe collection, processing, and tabulation of the data in aneffort to minimize its influence.
A potential source of bias in the estimates is due toimputing data for nonrespondents and for data which failedit. (Imputation is the process of replacing a missingvalue by a predicted value obtained from an appropriatemodel for nonresponse.) Across all kinds of business com-bined, in any given month imputed data amounts to about26 percent of the total retail sales estimate and about 32percent of the end-of-month inventory estimate. For theannual survey, imputed data amounts to about 9 percentof the total retail sales estimate, 11 percent of the totalend-of-year inventory estimate, 25 percent of the totalpurchases estimate, and 11 percent of the total accountsreceivable estimate.
A-5
U.S. Census Bureau
Table A-1. Estimated Coefficients of Variation for Monthly Retail Sales and End-of-MonthInventories by Kinds of Business: 1997
[Coefficients are percentages—based on sales and inventories estimates not adjusted for seasonal variations, holiday, or trading-day differences]
(NA) Not available.1Includes data for leased departments operated within department stores. Data for this line not included in broader kind-of-business totals.2Includes establishments primarily selling merchandise through coin-operated vending machines, by house-to-house canvass, or from mail order.3GAF represents stores which specialize in department store types of merchandise (general merchandise, apparel, furniture, miscellaneous shopping goods stores).
Note: The ranges and medians of the coefficients of variation shown above are based on sales and inventories estimates for the 12 most recent data months. Coefficients of variation for historicalestimates are available in prior publications.
A-6
U.S. Census Bureau
Table A-2. Estimated Coefficients of Variation for Annual Sales, End-of-Year Inventories, and Purchases: 1997[In percent]
(NP) Not published.1GAF represents stores which specialize in department store types of merchandise (general merchandise, apparel, furniture, miscellaneous shopping goods stores).
Note: The estimated coefficients of variation are based on 1997 Annual Retail Trade Survey data, adjusted to preliminary results of the 1997 Census of Retail Trade. Coefficients of variation forhistorical estimates are available in prior publications. Further detail for year-end inventories are available upon request.
Table A-3. Estimated Coefficients of Variation for Accounts Receivable Dollar VolumeEstimates: 1997
*Measure of sampling variability are not shown because the corresponding total estimates are not published in Table 6.
Note: The estimates of sampling variability shown above are based on estimates derived from the 1997 Annual Retail Trade Survey. Coefficients of variation for historical estimates are available inprior publications.
A-7
U.S. Census Bureau
Appendix B.Kind-of-Business Classifications
GENERAL DESCRIPTION
Retail trade SIC Major Groups 52-59, includes establish-ments primarily engaged in selling merchandise for per-sonal or household consumption and rendering servicesincidental to the sale of goods.
The kinds of business shown in this report parallelbroad classifications defined in the 1987 edition of theStandard Industrial Classification (SIC) manual. Descrip-tions of each kind of business follow. Data for some kindsof business are not shown separately but are included inU.S. and group totals.
DURABLE GOODS
Building Materials Group Stores(SIC Major Group 52)
This major group includes retail establishments primar-ily engaged in selling lumber and other building materials;paint, glass, and wallpaper; hardware; nursery stock; lawnand garden supplies; and mobile homes even if sales tocontractors account for the larger proportion of totalsales. These establishments are known as ‘‘retail’’ in thetrade. Establishments primarily selling these products butnot selling to the general public are classified in wholesaletrade.
Lumber and Other Building Materials Dealers (SIC521).1 Establishments engaged in selling primarily lum-ber and a general line of building materials to the generalpublic. The lumber which they sell may include rough anddressed lumber, flooring, molding, doors, sashes, frames,and other millwork. The building materials may includeroofing, siding, shingles, wallboard, paint, brick, tile,cement, sand, gravel, and other building materials andsupplies. Hardware is often an important line of retail lum-ber and building materials dealers. Establishments knownas ‘‘home centers’’ are included here.
Paint, Glass, and Wallpaper Stores (SIC 523).1 Estab-lishments engaged in selling primarily paint, glass, andwallpaper, or any combination of these lines, to the gen-eral public.
Hardware Stores (SIC 525). Establishments primarilyselling a number of basic hardware lines such as tools,builders’ hardware, paint and glass, housewares andhousehold appliances, cutlery, and roofing materials, noone of which accounts for 50 percent or more of the salesof the establishments.
Retail Nurseries, Lawn and Garden Supply Stores(SIC 526).1 Establishments primarily engaged in sellingtrees, shrubs, other plants, seeds, bulbs, mulches, soilconditioners, fertilizers, pesticides, garden tools, andother garden supplies to the general public. These estab-lishments primarily sell products purchased from others,but may sell some plants which they grow themselves.
Mobile Home Dealers (SIC 527).1 Establishments pri-marily engaged in the retail sale of new and used mobilehomes, including parts and accessories.
Automotive Dealers(SIC Major Group 55 ex. 554)
This major group includes retail dealers selling new andused automobiles, boats, recreational trailers, and motor-cycles; and those selling new automobile parts and acces-sories. Automobile repair shops maintained by the estab-lishments engaged in the sale of new automobiles are alsoincluded.
Motor Vehicle Dealers (New and Used) (SIC 551).Establishments primarily engaged in the sale of new auto-mobiles or new and used foreign or domestic automo-biles. These establishments frequently maintain repairdepartments and used car lots and carry stocks of replace-ment parts, tires, batteries, and automotive accessories.Used car lots and repair departments operated by fran-chised new passenger car dealers are not considered sepa-rate locations.
Motor Vehicle Dealers (Used Cars) Only (SIC 552).1
Establishments primarily selling used cars and not holdinga franchise for the sale of new cars.
Auto and Home Supply Stores (SIC 553). Establish-ments primarily engaged in the retail sale of automobiletires, batteries, and other automobile parts and accesso-ries. These establishments frequently sell additional linesof merchandise such as household appliances, radios andtelevision sets, sporting goods, housewares, and hard-ware.
Boat Dealers (SIC 555).1 Establishments primarilyengaged in the retail sale of new and used motorboatsand other watercraft, including parts, accessories, marinesupplies, and outboard motors.
B-1
U.S. Census Bureau
Recreational Vehicle Dealers (SIC 556).1 Establish-ments primarily engaged in the retail sale of new and usedrecreational trailers, campers (pickup coaches), and othertrailers for passenger automobiles, and motor homes,including parts and accessories.
Motorcycle Dealers (SIC 557).1 Establishments primar-ily engaged in the retail sale of new and used motorcyclesand motor scooters, including parts and accessories.
Automotive Dealers, Not Elsewhere Classified (SIC559).1 Establishments primarily engaged in the retail saleof new and used automotive vehicles, such as snowmo-biles, dunebuggies, gocarts, aircraft, utility trailers, andnew automotive equipment and supplies, not elsewhereclassified.
Furniture Group Stores (SIC Major Group 57)
This major group includes retail stores selling goodsused for furnishing the home, such as furniture, floor cov-erings, draperies, glassware and china, domestic stoves,refrigerators, and other household appliances and elec-tronics.
Furniture Stores (SIC 5712). Establishments primarilyselling household furniture, beds, mattresses, springs, andother sleep equipment. Also included in this classificationare establishments selling household appliances, phono-graphs, radio and TV sets, and floor coverings, providedthe receipts from sales of furniture and sleep equipmentexceed those from sales of other merchandise.
Floor Coverings Stores (SIC 5713). Establishmentsprimarily engaged in the retail sale of floor coverings andrelated products.
Drapery, Curtain, and Upholstery Stores (SIC 5714).1
Establishments primarily selling draperies, curtains, slip-covers, and upholstery materials. Establishments primarilyselling custom made draperies and slipcovers for house-hold use also are included.
Miscellaneous Home Furnishing Stores (SIC 5719).1
Establishments primarily engaged in the retail sale ofmiscellaneous home furnishings, such as china, glassware,and metalware for kitchen and table use, bedding andlinen, brooms, brushes, lamps and shades, mirrors andpictures, venetian blinds, window shades, and fireplaceaccessories.
Household Appliance Stores (SIC 5722). Establish-ments primarily engaged in the retail sale of electric andgas refrigerators, stoves, and other household appliances,such as electric irons, percolators, hot plates, and vacuumcleaners. Many such stores also sell radio and television
sets. Retail stores operated by public utility companiesand primarily engaged in the sale of electric and gas appli-ances for household use are classified here.
Radio, Television, and Electronics Stores (SIC 5731).Establishments primarily engaged in the retail sale andinstallation of radios, televisions, record players, soundreproducing equipment and home computers. Such estab-lishments also may sell additional lines, such as house-hold appliances, musical instruments, or records.
Computer and Computer Software Stores (SIC 5734).Establishments primarily engaged in the retail sale of com-puters, computer peripheral equipment, and software.
Music Stores (SIC 5735+5736).1 Establishments pri-marily engaged in the retail sale of musical instruments,phonograph records and albums, sheet music, and similarmusical supplies.
This includes all retail durable goods stores not else-where classified.
Used Merchandise Stores (SIC 593).1 This industryincludes stores primarily engaged in the retail sale of usedmerchandise, antiques, and secondhand goods such asclothing and shoes; furniture; books and rare manuscripts;musical instruments; office furniture, phonographs andphonograph records; and store fixtures and equipment.This industry also includes pawnshops.
Sporting Goods Stores and Bicycle Shops (SIC 5941).Establishments primarily selling a general or a specializedline of sporting goods and equipment for hunting, camp-ing, fishing, skiing, riding, tennis, golf, and other sports;bicycles, bicycle parts and accessories; and gymnasiumand playground equipment.
Book Stores (SIC 5942). Establishments primarily sell-ing new books and periodicals. Stationery and relateditems may also be sold.
Jewelry Stores (SIC 5944). Establishments primarilyengaged in the retail sale of any combination of the linesof jewelry, such as diamonds and other precious stones;rings, bracelets, and brooches; sterling and plated silver-ware; and watches and clocks.
Hobby, Toy, and Game Shops (SIC 5945).1 Establish-ments primarily engaged in the retail sale of toys, games,crafts, hobby kits and supplies.
Camera and Photographic Supply Stores (SIC 5946).1
Establishments primarily engaged in the retail sale of cam-eras, film, and other photographic supplies and equip-ment. Excluded are establishments primarily engaged infinishing films.
B-2
U.S. Census Bureau
Gift, Novelty, and Souvenir Shops (SIC 5947).1 Estab-lishments primarily engaged in the retail sale of combinedlines of gifts and novelty merchandise, souvenirs, andmiscellaneous small art goods such as greeting cards andholiday decorations.
Luggage and Leather Goods Stores (SIC 5948).1
Establishments primarily engaged in the retail sale of lug-gage, trunks, and leather goods.
Optical Goods Stores (SIC 5995).1 Establishments pri-marily engaged in the retail sale of eyeglasses and relatedoptical goods. Excluded are establishments whose receiptsare primarily from examining eyes and prescribing eye-glasses or contact lenses.
Miscellaneous Retail Stores Not Elsewhere Classi-fied (SIC 5999).1 Establishments primarily engaged inthe retail sale of specialized lines of merchandise such ascollectors’ items and supplies, artists’ supplies, orthopedicand artificial limbs, drafting materials, cosmetics, tele-phones and communication devices, pets, religious goods,hearing aids, monuments and tombstones, picture frames,swimming polls, and other merchandise not elsewhereclassified.
NONDURABLE GOODS
General Merchandise Group Stores (SIC MajorGroup 53)
This major group includes retail stores which sell anumber of lines of merchandise, such as dry goods,apparel and accessories, furniture and home furnishings,small wares, hardware, and food. The stores included inthis group are known as department stores, variety stores,general merchandise stores, general stores, etc.
Department Stores (SIC 531). Establishments normallyemploying 50 people or more, having sales of apparel andsoft goods combined amounting to 20 percent or more oftotal sales, and selling each of the following lines of mer-chandise:
• Furniture, home furnishings, appliances, radio and TVsets.
• A general line of apparel for the family.
• Household linens and dry goods.
To qualify as a department store, sales of each of thelines listed above must be less than 80 percent of totalstore sales. An establishment with total sales of $10 mil-lion or more is classified as a department store even ifsales of one of the merchandise lines listed above exceedthe maximum percent of total sales, provided that thecombined sales of the other two groups are $1 million ormore. Department store sales have been separately pre-sented for the following classifications:
Department stores (incl. leased depts.), andDepartment stores (excl. leased depts.).
Department stores (incl. leased depts.) are also sub-categorized as shown below:
Conventional Department Stores (SIC 531 part).Establishments which satisfy the criteria of a departmentstore and:
• Usually provide check-out service and customer assis-tance (salespersons) within each department.
• May have a catalog order desk.
• Are not affiliated with a company which operates similarestablishments on a national basis.
These stores often sell:
• Soft goods and hard goods which are primarily nation-ally advertised brands.
• Appliances which are serviced by another company.
• Limited lines of merchandise through seasonal or spe-cial catalogs.
Discount Department Stores (SIC 531 part). Estab-lishments which satisfy the criteria of a department storeand usually:
• Convey the image of a high-volume, fast turnover outletselling a variety of merchandise for less than conven-tional prices.
• Provide centralized check-out service.
• Do not provide customer assistance within store depart-ments. Merchandise is normally sold through self-service with minimal assistance provided in any depart-ment.
• Do not have a catalog order desk.
These stores often sell:
• Soft goods which are usually their own corporatebrands or are unbranded.
• Hard goods which are primarily nationally advertisedbrands.
• Appliances which are serviced by another company.
National Chain Department Stores (SIC 531 part).Establishments which satisfy the criteria of a departmentstore and:
• Usually provide check-out service and customer assis-tance (salespersons) within each department.
• May have a catalog order desk.
• Are affiliated with a company which operates similarestablishments on a national basis.
B-3
U.S. Census Bureau
These stores often sell:
• Soft goods and hard goods which are their own corpo-rate brands or are unbranded.
• Appliances which are serviced by their own company.
Variety Stores (SIC 533). Establishments primarilyengaged in the retail sale of a variety of merchandise inthe low and popular price ranges. Sales usually are madeon a cash-and-carry basis with the open selling method ofdisplay and customer selection of merchandise. Thesestores generally do not carry a complete line of merchan-dise, are not departmentalized, do not carry their owncharge service, and do not deliver merchandise.
Miscellaneous General Merchandise Stores (SIC539). Establishments primarily engaged in the retail saleof a general line of apparel, dry goods, hardware, home-wares or home furnishings, groceries, and other lines inlimited amounts. Stores selling commodities covered inthe definition for department stores, but normally havingless than 50 employees, and stores usually known as cata-log showrooms, country general stores, or warehouseclubs are included here. Also included in this classificationare establishments whose sales of apparel or of furnitureand home furnishings exceed half of their total sales pro-viding that sales of the smaller of the two lines in combi-nation with dry goods and household linens account for20 percent or more of total sales.
Food Group Stores (SIC Major Group 54)
This major group includes retail stores primarilyengaged in selling food for home preparation and con-sumption.
Grocery Stores (SIC 541). Establishments primarily sell-ing (1) a wide variety of canned or frozen foods, such asvegetables, fruits, and soups; (2) dry groceries, eitherpackaged or in bulk, such as tea, coffee, cocoa, driedfruits, spices, sugar, flour, and crackers; and (3) other pro-cessed foods and nonedible grocery items. In addition,these establishments often sell smoked and preparedmeats, fresh fish and poultry, fresh vegetables and fruits,and fresh or frozen meats.
Establishments commonly known as supermarkets,food stores, convenience stores, and delicatessens areincluded in this classification if receipts from sales of gro-ceries and food items for off-premise preparation and con-sumption are 50 percent or more of total sales.
Meat and Fish (Seafood) Markets, Including FreezerProvisioners (SIC 542). Establishments primarilyengaged in the retail sale of fresh, frozen, or cured meats,fish, shellfish, and other seafood. Frequently these estab-lishments also sell poultry, dairy products, eggs, somegroceries, and items commonly used in preparing seafoodor consumed with seafood.
Fruit Stores and Vegetable Markets (SIC 543).1
Establishments primarily selling fresh fruits and fresh veg-etables. These establishments frequently carry a limitedline of grocery items. Roadside stands of farmers sellingonly their own produce are not included.
Candy, Nut, and Confectionery Stores (SIC 544).1
Establishments primarily selling candy, nuts, sweetmeats,and other confections. A soda fountain or lunch counter isfrequently operated in these stores. Also included arecandy and popcorn stands located in motion picture the-aters.
Dairy Products Stores (SIC 545).1 Establishments pri-marily engaged in the retail sale of dairy products, such asmilk, cream, butter, cheese, and related products, to over-the-counter customers.
Retail Bakeries (SIC 546). Establishments primarilyengaged in the over-the-counter retail sale of bakery prod-ucts such as bread, cakes, pies, or cookies, all or some ofwhich may be baked on the premises.
Other Miscellaneous Food Stores (SIC 549).1 Estab-lishments primarily engaged in the retail sale of special-ized foods, not elsewhere classified, such as eggs andpoultry, health foods, spices, herbs, coffee, and tea. Thepoultry stores may sell live poultry, slaughter and cleanpoultry for their own account and sell dressed fowls, orsell fowl cleaned and dressed by others.
Gasoline Service Stations (SIC 554)
Establishments primarily selling gasoline and automo-tive lubricants. Usually these establishments also sell tires,batteries, and accessories, and perform minor repair workand services. Establishments called garages but derivingmore than half of their receipts from the sale of gasolineand automotive lubricants and establishments known as‘‘truck stops’’ which are primarily engaged in selling dieselfuel to truckers are also included. Convenience food storesderiving more than 50 percent of their sales from the saleof gasoline are classified as gasoline service stations.
Apparel and Accessory Stores(SIC Major Group 56)
Establishments in this major group are primarilyengaged in selling clothing of all kinds and related articlesfor personal wear and adornment.
Men’s and Boys’ Clothing Stores (SIC 561). Establish-ments primarily selling men’s and boys’ ready-to-wearclothing and furnishings.
Women’s Accessory and Specialty Stores (SIC 563).1
Establishments primarily selling a specialized line of wom-en’s and girls’ apparel, such as furs, sportswear, beach-wear, blouses, hosiery, millinery, foundation garments, lin-gerie, robes, and other intimate wear.
Children’s and Infants’ Wear Stores (SIC 564).1
Establishments primarily engaged in the retail sale of chil-dren’s and infants’ clothing, furnishings, and accessories.Such establishments may specialize in either children’s orinfants’ wear or they may sell a combination of children’sand infants’ wear.
Family Clothing Stores (SIC 565). Establishments pri-marily selling clothing, furnishings, and accessories formen, women, and children, without specializing in anyone line.
Shoe Stores (SIC 566). Establishments primarilyengaged in the retail sale of any one line, or a combinationof the lines, of men’s, women’s, and children’s footwear.These establishments frequently carry accessory linessuch as hosiery, gloves, and handbags.
Miscellaneous Apparel and Accessory Stores (SIC569).1 Establishments primarily engaged in the retail saleof specialized lines of apparel and accessories not else-where classified, such as uniforms, bathing suits, rain-coats, riding apparel, sports apparel, umbrellas, wigs, andtoupees. This industry also includes custom tailors primar-ily engaged in making and selling men’s and women’sclothing (except fur apparel SIC 568) to individual order.
Eating and Drinking Places(SIC Major Group 58)
Establishments in this major group are primarilyengaged in selling prepared foods and drinks for con-sumption on or near the premises and lunch counters andrefreshment stands selling prepared foods and drinks forimmediate consumption.
Restaurants, Lunchrooms, Cafeterias, and ContractFeeding (SIC 5812 part).
Restaurants and Lunchrooms.1 Establishments engagedin serving prepared food and beverages selected by thepatron from a full menu. Waiter or waitress service isprovided and the establishment has seating facilities forat least 15 patrons. These establishments often servealcoholic beverages, but receipts from the sale ofalcoholic beverages may not exceed the receipts fromprepared food.
Cafeterias.1 Establishments engaged in serving pre-pared food and beverages primarily through the use ofa cafeteria line where customers serve themselves fromdisplayed selections. Some limited waiter or waitressservice may be provided. Table and/or booth seatingfacilities are usually provided.
Contract Feeding.1 Establishments primarily engaged inproviding food service under contract to another com-pany; hospital; or governmental, penal, or educationalinstitution. The facilities and personnel of these estab-lishments may be provided by the contracting com-pany, institutions, etc., however, the management isalways supplied by the contractor.
Social Caterers (SIC 5812 part).1 Establishments pri-marily engaged in serving prepared food and beveragesfor weddings, banquets, etc., at a hall or similar placerather than at a fixed business location. Such establish-ments also may arrange for some entertainment but thisshould be a minor part of the business.
Refreshment Places, Ice Cream and Frozen CustardStands, (SIC 5812 part).
Refreshment Places.1 Establishments primarily sellinglimited lines of refreshments and prepared food items.Included in this group are establishments which pre-pare refreshment items such as pizza, barbecuedchicken, and hamburgers for consumption either on ornear the premises or for ‘‘take-home’’ consumption.
Ice Cream and Frozen Custard Stands.1 Establishmentsprimarily engaged in selling ice cream, frozen custard,or other frozen ices for consumption either on or nearthe premises. ‘‘Take-home’’ packages also may be pro-vided for ice cream sold in bulk.
Drinking Places (SIC 5813). Establishments primarilyengaged in the retail sale of drinks such as beer, ale, wine,liquor, and other alcoholic beverages for consumption onthe premises. Prepared foods are frequently sold at theseestablishments, but receipts from the sale of preparedfoods may not exceed receipts from sales of alcoholic bev-erages.
Drug Stores and Proprietary Stores (SIC 591)
Establishments engaged in the retail sale of prescriptiondrugs and patent medicines and which may carry a num-ber of related lines such as cosmetics, toiletries, tobacco,and novelty merchandise. Included are drug stores whichalso have a soda fountain or lunch counter. These storesare included on the basis of their usual trade designationrather than on the more strict interpretation of commodi-ties handled.
Liquor Stores (SIC 592)
Establishments primarily selling packaged alcoholicbeverages, such as ale, beer, wine, and whiskey, for con-sumption off the premises. Liquor stores operated byStates, counties, and municipalities are included.
Miscellaneous Nondurable Goods Stores(SIC Major Group 59, part)
This includes all retail nondurable goods stores notelsewhere classified.
B-5
U.S. Census Bureau
Miscellaneous Shopping Goods Stores (SIC 594).Establishments engaged in the retail sale of one or morelines of merchandise similar to that found indepartment stores.
Stationery Stores (SIC 5943).1 Establishments primar-ily engaged in the retail sale of stationery, such as paperand paper products (including printing and engraving),post cards, novelties, and school and office supplies(excluding office furniture and machines).
Sewing, Needlework, and Piece Goods Stores (SIC5949).1 Establishments primarily engaged in the retailsale of piece goods, notions, sewing and knitting supplies,fabrics, patterns, and other needlework accessories.
Mail-Order Houses (Department Store Merchandise)(SIC 5961 part). Establishments with normally 50 ormore employees primarily engaged in the retail sale bycatalog and mail order of a general line of merchandisesimilar to that sold by department stores.
Other Mail-Order Houses (SIC 5961 part). Establish-ments primarily engaged in the retail sale of a specializedor limited line of merchandise such as food, automotivemerchandise, apparel, books, stationery, etc., by catalogand mail order.
Establishments primarily engaged in the retail sale ofproducts by means of automatic merchandising units(vending machines) which are generally located on thepremises of other businesses. Those products includecandy, nut, and confectionery; milk and ice cream; otherbeverages; and tobacco products.
Direct-Selling Operations (SIC 5963).1 Establishmentsprimarily selling merchandise by house-to-house canvass,by party plan, by telephone, or from a truck. The merchan-dise includes building materials, hardware, and gardensupplies; general merchandise, milk; other foods; appareland accessories; furniture, home furnishings, and equip-ment; mobile food service; and books and stationery.
Fuel Oil Dealers (SIC 5983).1 Establishments primarilyengaged in the retail sale of fuel oil.
Liquefied Petroleum Gas Dealers (SIC 5984).1 Estab-lishments primarily engaged in the retail sale of liquefiedpetroleum (LP) gas (bottled gas or in bulk.)
Fuel Dealers, Not Elsewhere Classified (SIC 5989)1
Establishments primarily engaged in the retail sale of coal,coke, charcoal, wood, or any combination of these lines.
Florists (SIC 5992).1 Establishments primarily engagedin the retail sale of cut flowers and growing plants. Green-houses and nurseries are included if receipts are primarilyfrom the sale of products not grown on the premises.
Cigar Stores and Stands (SIC 5993).1 Establishmentsprimarily engaged in the retail sale of cigars, cigarettes,tobacco, and smokers’ supplies. Many of the establish-ments included in this classification are operated asconcessions in places of amusement, railway stations, air-ports, and other public places.
News Dealers and Newsstands (SIC 5994).1 Establish-ments primarily engaged in the retail sale of newspapers,magazines, and other periodicals.
1Data for this kind of business are not shown separately but areincluded in U.S. and group totals.