1997 Economic Census:
Surveys of Minority- and Women-Owned Businesses

Introduction to the Economic Census
Purposes and uses of the Economic Census - Geographic area coding - Basis of reporting - Availabiltiy of additional data - Historical information - Sources for more information

Surveys of Minority- and Women-Owned Business
General - Survey methodology - Reliability of estimates - Census disclosure rules - Industry classifications - Geographic areas covered - Comparability of 1997 and 1992 data - Comparability of survey data with other economic census data - Dollar values - Special tabulations - Abbreviations and symbols


Introduction to the Economic Census



The economic census is the major source of facts about the structure and functioning of the Nation's economy. It provides essential information for government, business, industry, and the general public. Title 13 of the United States Code (Sections 131, 191, and 224) directs the Census Bureau to take the economic census every 5 years, covering years ending in 2 and 7.

The economic census furnishes an important part of the framework for such composite measures as the gross domestic product estimates, input/output measures, production and price indexes, and other statistical series that measure short-term changes in economic conditions. Specific uses of economic census data include the following:


Accurate and complete information on the physical location of each establishment is required to tabulate the census data for the states, metropolitan areas (MAs), counties, parishes, and corporate municipalities including cities, towns, villages, and boroughs. Respondents were required to report their physical location (street address, municipality, county, and state) if it differed from their mailing address. For establishments not surveyed by mail (and those single-establishment companies that did not provide acceptable information on physical location), location information from Internal Revenue Service tax forms is used as a basis for coding.


The economic census is conducted on an establishment basis. A company operating at more than one location is required to file a separate report for each store, factory, shop, or other location. Each establishment is assigned a separate industry classification based on its primary activity and not that of its parent company.


Reports in Print and Electronic Media

All results of the 1997 Economic Census are available on the Census Bureau Internet site (www.census.gov) and on compact discs (CD-ROM) for sale by the Census Bureau. Unlike previous censuses, only selected highlights are published in printed reports. For more information, including a description of electronic and printed reports being issued, see the Internet site, or write to U.S. Census Bureau, Washington, DC 20233-8300, or call Customer Services at 301-763-INFO(4636).


The economic census has been taken as an integrated program at 5-year intervals since 1967 and before that for 1954, 1958, and 1963. Prior to that time, individual components of the economic census were taken separately at varying intervals.

The economic census traces its beginnings to the 1810 Decennial Census, when questions on manufacturing were included with those for population. Coverage of economic activities was expanded for the 1840 Decennial Census and subsequent censuses to include mining and some commercial activities. The 1905 Manufactures Census was the first time a census was taken apart from the regular decennial population census. Censuses covering retail and wholesale trade and construction industries were added in 1930, as were some covering service trades in 1933. Censuses of construction, manufacturing, and the other business service censuses were suspended during World War II.

The 1954 Economic Census was the first census to be fully integrated: providing comparable census data across economic sectors, using consistent time periods, concepts, definitions, classifications, and reporting units. It was the first census to be taken by mail, using lists of firms provided by the administrative records of other Federal agencies. Since 1963, administrative records also have been used to provide basic statistics for very small firms, reducing or eliminating the need to send them census questionnaires.

The range of industries covered in the economic censuses expanded between 1967 and 1992. The census of construction industries began on a regular basis in 1967, and the scope of service industries, introduced in 1933, was broadened in 1967, 1977, and 1987. While a few transportation industries were covered as early as 1963, it was not until 1992 that the census broadened to include all of transportation, communications, and utilities. Also new for 1992 was coverage of financial, insurance, and real estate industries. With these additions, the economic census and the separate census of governments and census of agriculture collectively covered roughly 98 percent of all economic activity.

Printed statistical reports from the 1992 and earlier censuses provide historical figures for the study of long-term time series and are available in some large libraries. All of the census reports printed since 1967 are still available for sale on microfiche from the Census Bureau. CD-ROMs issued from the 1987 and 1992 Economic Censuses contain databases including nearly all data published in print, plus additional statistics, such as ZIP Code statistics, published only on CD- ROM.


More information about the scope, coverage, classification system, data items, and publications for each of the economic censuses and related surveys is published in the Guide to the 1997 Economic Census and Related Statistics at www.census.gov/epcd/www/guide.html. More information on the methodology, procedures, and history of the censuses will be published in the History of the 1997 Economic Census, on the web at www.census.gov/econ/www/history.h tml , and in print.

Surveys of Minority- And Women-owned Business Enterprises


The Survey of Minority-Owned Business Enterprises (SMOBE) is conducted in conjunction with the Survey of Women-Owned Business Enterprises (SWOBE). The SMOBE and SWOBE provide basic economic data on businesses owned by Blacks, persons of Alaska Native, American Indian, Asian, or Pacific Islander descent, persons of His- panic or Latin American ancestry, and women. These sur- veys are based on the entire firm rather than on individual locations of a firm. The published data cover number of firms, gross receipts, number of paid employees, and annual payroll. The data are presented by geographic area, industry, size of firm, and legal form of organization of firm.


All firms operating during 1997, except those classified as agricultural, are represented in these surveys. The lists of all firms (or universe) are compiled from a combination of business tax returns and data collected on other economic census reports. The Census Bureau obtains electronic files from the IRS for all companies filing Internal Revenue Service (IRS) Form 1040, Schedule C (individual proprietorship or self-employed person); 1065 (partnership); or any one of the 1120 corporation tax forms; and 941 (Employer's Quarterly Federal Tax Return). For businesses filing those forms, the IRS provided the Census Bureau with the following information: Name and address of the firm; employer identification number of the firm; Social Security numbers (SSN's) of the owners for filers of Form 1040 Schedule C and electronic filers of Forms 1065 and 1120S (subchapter S corporations); principal industrial activity code (see Comparability of 1992 and 1997 Data); dollar receipts; annual payroll; and legal form of organization.

For most firms with paid employees, the Census Bureau also collected employment, payroll, receipts, and kind of business information for each plant, store, or physical location during the 1997 Economic Census.

To design the sample several sources of information were used to identify the probability that a business was minority- or woman-owned. For all sole proprietorships and partnerships and corporations which filed an electronic tax return, the SSN of the owner(s) was provided to the Census Bureau by IRS. These SSNs were sent to Social Security Administration (SSA) which provided the race and sex codes that were indicated by the individuals on their applications for SSNs.

Persons applying for SSN's prior to 1981 could categorize their race as (a) White, (b) Black, or (c) Other. In 1981, the racial descriptions on social security applications were expanded to (a) Asian, Asian-American, or Pacific Islander, (b) Hispanic, (c) Black, (d) Northern American Indian or Alaskan Native, and (e) White.

Most persons who currently own businesses applied for their SSN's prior to 1981. Therefore, the majority of owners could be classified only as "White," "Black," or "Other" by use of SSA race codes.

For each SSN, the SSA also provided the Census Bureau with the individual's country of birth, current surname, original surname, mother's maiden surname, and father's surname. The Census Bureau has developed lists of American Indian, Asian, and Hispanic surnames based on research using prior survey data.

In addition to SSA data, several other sources were used to pre-identify businesses by race, ethnicity, and gender of owner(s) as potentially minority-owned:

Individual proprietors who reported they were of minority ancestry in the 1992 survey and were still active in 1997 were excluded from the mail canvass and the 1992 responses were used instead. See the section on SAMPLING below for a description of how cases were selected.

Sampling. There were six sampling frames used and every case was assigned to one of the following frames:

The SMOBE and SWOBE universe was stratified by state, industry, and the inferred race code. The Census Bureau selected large companies (based on volume of sales) 'with certainty'. All certainty cases were sure to be selected and represented only themselves (i.e., have a selection probability of one and a sampling weight of one). The certainty cutoffs varied by sampling stratum, and each stratum was sampled at varying rates, depending on the number of firms in a particular industry in a particular state. The sampling rate was lowest in states and industries with the greatest number of firms. A similar methodology was used to select a sample from the remaining universe: the purpose of this was to estimate the number of firms owned by persons of minority ancestry when no indication of minority ownership was found from any of the sources listed above.

A firm selected into the sample was mailed one of two questionnaires. The Census Bureau sent the MB-1 questionnaire to partnerships and corporations, or to sole proprietorships that submitted joint tax returns where, based on the administrative records, it was probable that the husband and wife were of different races. The businesses were asked to report the percentage of stock ownership of the business by gender, race and ethnicity. The MB-2 questionnaire was used for sole proprietors and self-employed individuals who were "single filers" or who filed joint tax returns where, based on administrative records, there was a low probability that the husband and wife were of different races. The businesses were asked to report the gender, race and ethnicity of the primary owner(s) of the business. The form included an equal male/female ownership option for the collection of business owners by gender. Copies of the MB-1 and MB-2 forms can be found at http://www.census.gov/epcd/www/ec97frm2.html.

Tabulation. In the 1997 SMOBE and SWOBE, minority/women ownership of a business was based on the race/ethnicity/gender of the person(s) owning majority interest in the business. Firms equally male-/female-owned were counted and tabulated as a separate category. The gender of sole proprietors and self-employed persons who were "single filers" was taken directly from administrative record data.

Businesses in which ownership was shared among minority and nonminority groups with no single racial/ethnic group having majority interest were tabulated as 50 percent minority-/50 percent nonminority-owned in the Company Summary publication and are excluded from the minority business counts.

It was possible for a firm to be classified in more than one minority group (one racial and one Hispanic ethnicity) and to be included in more than one minority report. For example, a firm could be included in both the Asian and Pacific Islander report and the Hispanic report if it were owned equally by two people, one of Japanese ancestry and one of Mexican ancestry. However, such a firm is counted only once at total levels in the Summary publication.

Data are tabulated by firm. A firm is a business organization consisting of one or more domestic locations in the same state and industry that were specified under common ownership or control. Firms with more than one location in the same industry within a state will be counted as one firm and the employment and annual payroll are summed for all those locations. Firms operating in more than one industry within a state are counted in each industry. Firms operating in more than one state are counted in each state where they do business.

Nonprofit and foreign-owned companies as well as those with publicly held stock whose ownership was indeterminate were tabulated separately.


The figures shown in these reports are, in part, estimated from a sample and will differ from the figures which would have been obtained from a complete census. Two types of possible errors are associated with estimates based on data from sample surveys: sampling errors and nonsampling errors. The accuracy of a survey result depends not only on the sampling errors and nonsampling errors measured, but also on the nonsampling errors not explicitly measured. For particular estimates, the total error may considerably exceed the measured errors. The following is a description of the sampling and nonsampling errors associated with the 1997 SMOBE/SWOBE.

Sampling variability. The particular sample used for these surveys is one of a large number of all possible samples of the same size that could have been selected using the same sample design. Estimates derived from the different samples would differ from each other. The relative standard error is a measure of the variability among the estimates from all possible samples. The estimated relative standard errors presented in the tables estimate the sampling variability, and thus measure the precision with which an estimate from the particular sample selected for this survey approximates the average result of all possible samples. Relative standard errors are applicable only to those published cells in which sample cases are tabulated. A relative standard error is an expression of the standard error as a percent of the quantity being estimated.

The sample estimate and an estimate of its relative standard error can be used to estimate the standard error and then construct interval estimates with a prescribed level of confidence that the interval includes the average results of all samples. To illustrate, if all possible samples were surveyed under essentially the same condition, and estimates calculated from each sample, then:

  1. Approximately 68 percent of the intervals from one standard error below the estimate to one standard error above the estimate would include the average value of all possible samples.
  2. Approximately 90 percent of the intervals from 1.6 standard errors below the estimate to 1.6 standard errors above the estimate would include the average value of all possible samples.

Thus, for a particular sample, one can say with specified confidence that the average of all possible samples is included in the constructed interval.

Example of a confidence interval. Suppose the estimated number of employer minority-owned firms in Agriculture Services is 7,139 and the estimated relative standard error is 2 percent. The estimated standard error is 7,139 x .02 = 143. An approximate 90-percent confidence interval is 7,139 plus or minus (1.6 x 143) or 7,139 plus or minus 228.8 or 6,910.2 to 7,367.8.

Nonsampling errors. All surveys and censuses are subject to nonsampling errors. Nonsampling errors are attributable to many sources: inability to obtain information for all cases in the universe, adjustments to the weights of respondents to compensate for non-respondents, imputation for missing data, data errors and biases, mistakes in recording or keying data, errors in collection or processing, and coverage problems.

Explicit measures of the effects of these nonsampling errors are not available. However, it is believed that most of the important operational and data errors were detected and corrected through an automated data edit designed to review the data for reasonableness and consistency. Quality control techniques were used to verify that operating procedures were carried out as specified.

Approximately 79 percent of the questionnaires were returned. For respondents which supplied some but not all of the requested information and for businesses selected with certainty which failed to respond, data were imputed using a procedure that ensured that the imputed data were consistent with the responses provided and all information available from administrative sources. Weights of the responding cases which were not selected with certainty were adjusted to compensate for the remainder of the cases which failed to respond. All of these procedures assume that (1) the administrative information is accurate and (2) non-respondents possess the same characteristics as respondents with similar demographic characteristics. Neither of these assumptions is exactly true.


The U.S. Census Bureau is prohibited by law from publishing any statistics that disclose information reported by individual companies. Publishing the number of firms, however, is not considered confidential by the Census Bureau rules. Therefore, the number of firms may be shown, even when other data must be withheld. The information and data obtained from the IRS, the SSA, and other sources are also treated as confidential and can be seen only by Census Bureau employees sworn to protect it from unwarranted disclosure.


The system published in the Standard Industrial Classification (SIC) Manual: 1987 (for sale by Superintendent of Documents,U.S.Government Printing Office, Washington,DC 20402. Stock No.041-001-00314-2) was used to classify firms for this report. Under this system, economic activities are coded on a numerical basis. Related activities are grouped at a primary level by industry division (i.e., manufacturing, construction, retail trade, etc.). Within each industry division, successive levels of detail of the economic activity are defined: major group (two-digit code); industry group (three-digit code); and industry (four-digit code). For example, one hierarchy in wholesale trade is:

SIC level SIC code Description
Industry division -- Wholesale trade
Major group 50 Durable goods
Industry group 506 Electrical goods
Industry 5064 Electrical appliances, television and radio sets

Data in this report are presented at the industry division or major group level, depending on the table. This study covers all industries in the SIC system except the following major industry groups:


These reports present data for the United States, each State and the District of Columbia; metropolitan areas (MA's), which include selected metropolitan statistical areas (MSA's), primary metropolitan statistical areas (PMSA's), and consolidated metropolitan statistical areas (CMSA's); counties; and places with 100 or more women- or minority-owned firms.

MSA's. The MSA's for which data are shown are among those defined by the Office of Management and Budget as of June 30, 1993. An MSA is an integrated economic and social unit with a population nucleus of at least 50,000 inhabitants (according to the 1990 Census of Population or a subsequent special census). Each MSA consists of one or more counties meeting standards of metropolitan character; in New England, cities and towns, rather than counties, are the component geographic units.

PMSA's. An MSA with a population of 1 million or more may be subdivided into PMSA's. A PMSA consists of a large urbanized county or a cluster of counties (cities and towns in New England) that demonstrates very strong internal economic and social links separate from the ties to other portions of its MSA. Where PMSA's are defined, the MSA of which they are component parts is redesignated a CMSA.

Counties. Data are shown for counties or county equivalents, defined as of January 1, 1997.

Places. Included in this publication are places of 2,500 inhabitants or more (according to the 1990 Census of Population or subsequent spe- cial census) incorporated as a city, borough, village, or town as of January 1,1997. In addition, statistics are presented for towns in the six New England states, New York, and Wisconsin, and townships in Michigan, Minnesota, New Jersey, and Pennsylvania with a 1990 census population of 10,000 or more.


The following changes were made in survey methodology in 1997 which affect comparability with past reports:

  1. The universe for the 1997 SMOBE and SWOBE was expanded to include all corporations. In 1992, only subchapter S corporations in addition to partnerships and sole proprietorships were included for SMOBE. A small sample of "C" corporations was included in SWOBE to provide estimates at the industry division level only.
  2. Previously, the SMOBE and SWOBE programs defined a business as an entity that filed a business income tax return. Businesses that had received one or more Employer Identification Numbers (EINs) may have filed tax returns under more than one EIN. In past surveys, each EIN was treated as a separate business. For 1997, all operations under the same ownership were defined as one company or business, irrespective of the number of the company's EINs. The definition eliminated the likelihood of surveying the same business owner more than once. This change had no effect on the employment and payroll data for these surveys, but slightly reduced the count for the number of businesses and the total receipts.
  3. In the past, ownership was based on the race/ethnicity/gender of the majority of the number of owners, without regard to the percentage of interest owned in the firm. In the 1997 SMOBE and SWOBE, minority/women ownership of a business was based on the race/ethnicity/gender of the person(s) owning majority interest in the business. Businesses with 50 percent or more women/minority owners have in the past been included in the women/minority business counts. In the 1997 SMOBE and SWOBE, businesses equally male-/female-owned were excluded from the women-owned business counts and tabulated as a separate category. Businesses in which ownership was shared among minority and nonminority groups with no single racial/ethnic group having majority interest were excluded from the minority business counts and tabulated as 50 percent minority-/50 percent nonminority-owned in the Company Summary publication.
  4. In the past, a small sample of the cases placed on the non-Hispanic white male frame was taken to estimate the number of firms owned by persons of minority ancestry when no indication of minority ownership was found from any of the sources used in stratification. The estimates from this sample were presented in a separate section of the Asian- and Pacific Islander-, American Indian- and Alaska Native-owned business publication and the Hispanic-owned business publication as an estimate of the undercount and were excluded from the detailed tables. In the 1997 SMOBE and SWOBE, the estimates from this part of the sample were included in the detailed tables of the publications.
  5. In prior surveys, the SMOBE universe of sole proprietorships was based on businesses filing an annual IRS Schedule C (Form 1040)''Profit or Loss From Business ''report. Employers were identified by first matching these Schedule C records to IRS Form 941 ''Employers Quarterly Federal Tax Return ''payroll reports. Not all Schedule C business income tax records that represented firms with employees could be matched. To estimate for sole proprietorship employers that were not matched to payroll tax returns, a portion of the sole proprietorship universe that reported cost of labor or wages as an expense item on the Schedule C was also considered employers. In addition, some cases with no cost of labor or wages reported but with large receipts had payroll imputed and were counted as employers.

    For 1997,the methodology was revised to identify sole proprietorship employers as only those that filed the IRS Form 941 payroll tax form. These businesses were included in the universe even if they could not be matched to a Schedule C. This resulted in a decrease in the estimate of firms with paid employees relative to the 1992 methodology.


The tables in each publication that show data for all U.S. firms are comparable to and include the minority-/women-owned firm data. Caution should be exercised in comparing data presented in this report with published or unpublished data from other reports of the 1997 Economic Census. Factors that affect comparability of data among censuses are industrial scope, business unit covered, receipts size, and coverage of nonemployers.

Industrial scope. Data in these reports are based on the 1987 SIC Manual (see Industry Classifications section).

Several SIC industries were in scope of the SMOBE and SWOBE but out of scope of the basic economic census. These industries included: Agriculture Services (07); Forestry (08); and Fishing, Hunting and Trapping (09).

Membership Organizations (86) were within the scope of the economic census but out of the scope of the SMOBE and SWOBE.

In this study, industry 6552 (land subdividers and developers, except cemeteries) is included with the construction industry group, unlike the Comparative Statistics, the only other 1997 Economic Census publication reported by SIC, which include it with industry 65 (real estate).

Business unit. Most of the economic censuses are conducted on an establishment or physical location basis; individually enumerating and tabulating each establishment owned by a firm. In the SMOBE and SWOBE, however, the whole firm was the primary unit for enumeration. Receipts size. Most of the economic censuses determine whether or not a firm may be counted as an active business according to its receipts size. The specified minimums vary by industry. For the SMOBE and SWOBE, a firm had only to file a business tax return with business receipts greater than $1,000 for it to be counted as an active business.

Receipts size. Most of the economic censuses determine whether or not a firm may be counted as an active business according to its receipts size. The specified minimums vary by industry. For a firm to be counted as an active business in the SMOBE and SWOBE, it had only to file a business tax return with business receipts greater than $1,000, except for the construction industry, which had no minimum business receipts cutoff.

Firms with no paid employees. Although nonemployer firms are included in these surveys, they are omitted from many of the economic census reports.


All dollar values presented in this report are expressed in current dollars, i.e., 1997 data are expressed in 1997 dollars and 1992 data in 1992 dollars. Consequently, when making comparisons to prior years, users should take into consideration the inflation that has occurred.


Special tabulations of data collected in the 1997 SMOBE and/or SWOBE may be obtained, depending on availability of time and personnel, on computer disk, or in tabular form. The data will be in summary form and subject to the same rules prohibiting disclosure of confidential information (including name, address, kind of business, or other data for individual firms) as are the regular publications.

Special tabulations are prepared on a cost basis. A request for a cost estimate, as well as exact specifications on the type and format of the data to be provided, should be directed to the Chief, Economic Census Branch, Census Bureau, Washington, DC 20233-6400. To discuss a special tabulation before submitting a request, call 301-457-3318.


The following abbreviations and symbols are used in the SMOBE and SWOBE publications:

D Withheld to avoid disclosing data for individual companies; data are included in higher-level totals.
(IC) Independent city.
(NA)Not available.
(X)Not applicable.
CDP Census designated place.
CMSAConsolidated metropolitan statistical area.
MAMetropolitan area.
MSA Metropolitan statistical area.
PMSAPrimary metropolitan statistical area.
SICStandard industrial classification.
n.e.c. Not elsewhere classified.
n.s.k.Not specified by kind.


Many persons participated in the various activities of the 1997 Surveys of Minority- and Women-Owned Business Enterprises.

The Company Statistics Division prepared this report. Ruth A. Runyan, Assistant Chief for Surveys and Programs, was responsible for the overall planning, management, and coordination. Planning and implementation were under the direction of Eddie J. Salyers, Chief, Economic Census Branch, assisted by Valerie C. Strang, Section Chief. Primary staff assistance was provided by Melody Atkinson, Ahmad Bakhshi, Lori Bowan, Trey Cole, Elaine Emanuel, Barbara Esworthy, Mary Frauenfelder, Geoffrey Hill, James Jarzabkowski, Durwin Knutson, James McFarland, Sungsoo Oh, and John Seabold.

General direction for statistical methodology was provided by Ruth Detlefsen, Assistant Chief for Research and Methodology (until October 1998); Carol V. Caldwell, her successor; and Mark Sands, Chief, Statistical Research and Methods Branch. Richard Moore developed the sample design, imputation, estimation and variance methodology, with assistance from Lieu Galvin, David Harville, and Steven Klement.

The Economic Statistical Methods and Programming Division, Charles P. Pautler, Jr., Chief, assisted by Martin S. Harahush, Assistant Chief for Quinquennial Programs, developed and implemented computer processing systems. Steven G. McCraith, Chief, Quinquennial Surveys Branch, and Gary Sweetland, Chief, Mailout and Data Collection Branch, supervised the preparation of computer programs. Thaddeus S. Hess, Joseph F. Keehan, and Daniel A. Vacca were assigned primary programming responsibilities with assistance from John D. Bedwell, Edward F. Carr, Jr., Diane Conley, Steve Jarvis, Sheila Koonce, John E. McCormick, Robert A. Penrod, Linda Raley, Jerry W. Richards, and Michele Shelby.

The staff of the Decennial Systems and Contract Management Office, Alan Berlinger, Chief, were responsible for overseeing the electronic data capture operation.

Kim D. Ottenstein, Bernadette J. Gayle, Helen M. Curtis, and Laurene V. Qualls of the Administrative and Customer Services Division, Walter C. Odom, Chief, provided publications and printing management, graphics design and composition, and editorial review for print and electronic media. General direction and production management were provided by Michael G. Garland, Assistant Chief, and Gary J. Lauffer, Chief, Publications Services Branch.

The staff of the National Processing Center, Judith N. Petty, Chief, performed mailout preparation and receipt operations, clerical and analytical review activities, and data entry.

The Computer Services Division, Debra D. Williams, Chief, performed the computer processing.

The overall planning and review of the census operations were performed by the staff of the office of the Assistant Director for Economic and Agriculture Censuses.

This survey depended upon the cooperation of the staffs of the Internal Revenue Service and the Social Security Administration. Thanks are due to Daniel F. Skelly, Director, Statistics of Income Division, Karen Cys, Nick Greenia, and Ken Szeflinski of the Internal Revenue Service; Frank Thomas of the Social Security Administration; and Roberta Custard, Paul Hanczyark, and Edward D. Walker of the Census Bureau, who coordinated activities with the agencies.

If you have any questions concerning the statistics in this report, call (301) 457-3316.