Skip Header

Component ID: #ti1058327173

Survey Design

Target population

The target population of the Economic Census consists of all establishments that:

  • Are in operation for at least some part of 2017
  • Are located in one of the 50 states, associated offshore areas, or the District of Columbia.
  • Have paid employees
  • Classified in one of eighteen in-scope sectors defined by the North American Industry Classification System (NAICS). 

Component ID: #ti1681923509

The 2017 Economic Census covers the following NAICS sectors of the U.S. economy:

  • 21 – Mining, Quarrying, and Oil and Gas Extraction
  • 22 – Utilities
  • 23 – Construction
  • 31-33 – Manufacturing
  • 42 – Wholesale Trade
  • 44-45 – Retail Trade
  • 48-49 – Transportation and Warehousing
  • 51 – Information
  • 52 – Finance and Insurance
  • 53 – Real Estate and Rental and Leasing
  • 54 – Professional, Scientific and Technical Services
  • 55 – Management of Companies and Enterprises
  • 56 – Administrative and Support and Waste Management and Remediation Services
  • 61 – Educational Services
  • 62 – Health Care and Social Assistance
  • 71 – Arts, Entertainment, and Recreation
  • 72 – Accommodation and Food Services
  • 81 – Other Services (Except Public Administration)

More detailed descriptions of these sectors can be found at the Census Bureau NAICS webpage,

Component ID: #ti1264422518

The following industries (NAICS) are not covered by the 2017 Economic Census:

  • 11 – Agriculture, Forestry, Fishing and Hunting
  • 482 – Rail Transportation
  • 491 – Postal Service
  • 525 – Funds, Trusts, and Other Financial Vehicles
  • 6111 – Elementary and Secondary Schools
  • 6112 – Junior Colleges
  • 6113 – Colleges, Universities, and Professional Schools
  • 8131 – Religious Organizations
  • 81393 – Labor Unions and Similar Labor Organizations
  • 81394 – Political Organizations
  • 814 – Private Households
  • 92 – Public Administration

Component ID: #ti334155173

Sampling frame

The Economic Census selects establishments for its sample from a frame obtained from the U.S. Census Bureau’s Business Register.  The Business Register contains information on the physical location of establishments, as well as payroll, employment, receipts (value of shipments), and industry classification data obtained from prior censuses and surveys or obtained from the administrative records of the Internal Revenue Service (IRS) and Social Security Administration (SSA) under special arrangements which safeguard the confidentiality of both tax and census records.  Information from the Bureau of Labor Statistics on industry classifications are also used to supplement the classification information from the IRS and SSA.

To improve the quality of the frame, in 2016, the Census Bureau conducted a classification survey to update the NAICS industry classification of those establishments on the Business Register that could not be adequately classified into a NAICS industry code.

To be included on the sampling frame, an establishment was required to satisfy the following conditions:

  • It must be classified into an in-scope NAICS industry;
  • It must be an active establishment of a multi-establishment firm, or it must be a single-establishment firm with at least one quarter of 2017 administrative payroll; and
  • It must be located in one of the 50 states, associated offshore areas, or the District of Columbia. 

Component ID: #ti2113283650

Sampling unit

The basic sampling unit is the establishment.

Component ID: #ti1747893733

The sample component of the 2017 Economic Census consists of four parts:

  • All active operating establishments of multi-establishment firms (“Multi-Estab”),
  • All single-establishment firms whose annualized payroll was above an industry-specific payroll cutoff (“Large Certainty”),
  • Certain single-establishment firms taken with certainty for some specific reason (“Special Certainty”), or
  • A stratified sample selected from the remaining, generally smaller, single-establishment firms (“Non-Certainty”).

In order to reduce respondent burden and costs, the Census Bureau did not require all establishments to make a Census report.  For tabulations of basic data items (sales, payroll, employment, etc.) we use the administrative data of these, generally smaller, single establishment firms.  This non-sampled component consists of:

  • Small single-establishment firms on the sampling frame but not selected into the Census sample.  However, some of these establishments received an Economic Census request for industry classification information so that their administrative data could be accurately tabulated in the correct industry.  The column, “Classification Reqd” in the table below indicates the number of single-establishment firms that were required to provide this classification information.  The remaining single establishment firms were not contacted for the Economic Census (“Not Mailed” in the table below).
  • Single-establishment firms with no classification information.  There were approximately 20,000 of these firms and while they were not included in the sample, they were asked to provide industry classification information only.
  • Any single-establishment firms that started business operations so late in 2017 (mostly in the 4th quarter) that their 2017 administrative payroll was not received by the Census Bureau until after the 2017 sample was selected – roughly 260,000 firms.

As mentioned above, for the non-sampled component, the Economic Census uses imputation and administrative data from other federal agencies in lieu of requiring a response from these single-establishment firms.

The table below shows the sizes of each of the sample components and the first non-sampled component.  The last column shows the total number of single establishment firms that were on the sampling frame.

Component ID: #ti1125743832

Selection procedures differ between multi- and single-establishment firms.

Multi-Establishment Firms

Any firm with more than one active establishment is included in the Economic Census with certainty and is generally expected to report for all of its establishments.  Each establishment is included with certainty and assigned a sample weight of 1.

Establishment Reporting Units

In most industries, multi-establishment firms are required to complete an industry-specific questionnaire for each of the establishments in their firm.

Alternative Reporting Units (ARU) for Selected Industries

In some industries, firms have difficulty reporting revenue and related data for each of their business locations (establishments).  However, they can provide firm-level industry totals with relative ease and they can report separate payroll and employment information for each business location within the industry.  The table below shows the industries for which an alternative questionnaire was used and the expected number of affected firms.

If a firm had more than two establishments in an industry listed below, the firm received one questionnaire for each of those industries.  Each questionnaire requested consolidated, firm-level data for revenue and related measures covering the firm’s nationwide operations.  A supplementary questionnaire enumerated the firm’s establishments in the industry and requested payroll and employment information for each of them.

Component ID: #ti2123938372

Alternative Reporting Unit Industries and Counts


NAICS Description


Number of Firms

Number of Affected Establishments


Drilling Oil and Gas Wells 




Support Activities for Oil and Gas Operations 




Support Activities for Coal Mining




Support Activities for Metal Mining




Support Activities for Nonmetallic Minerals (except Fuels) Mining




Hydroelectric power generation




Fossil fuel electric power generation




Nuclear electric power generation 




Solar electric power generation 




Wind electric power generation 




Geothermal electric power generation 




Biomass electric power generation 




Other electric power generation 




Electric bulk power transmission and control 




Electric power distribution




Natural gas distribution




Water supply and irrigation systems 




Sewage treatment facilities 




Steam and air-conditioning supply 




Motion picture and video production 




Motion picture and video distribution 




Integrated record production/distribution 




Wired telecommunications distribution




Cable and other program distribution








Cellular and other wireless telecommunications




Commercial banking




Savings institutions 




Direct life insurance carriers 




Direct health and medical insurance carriers 




Direct property and casualty insurance carriers 




Direct title insurance carriers 




Other direct insurance (except life, health, and medical) carriers




Reinsurance carriers






Component ID: #ti1090027797

Single Establishment Firms with 2017 Payroll

The sample design for single-establishment firms began with a study of the potential respondent universe.  This study produced a set of industry-specific payroll cutoffs that were used to distinguish large single-establishment firms from small ones within each industry.  In general, these cutoffs were chosen so that the sum of the payroll of the multi-establishment firms plus the payroll of the single-establishment firms above the cutoff equaled 75-95% of the total payroll in an industry, though there were exceptions.  In the hypothetical example below, a payroll cutoff of $229,000 for an industry will result in 80% of total industry payroll (32.6% + 47.4%) being contained within the 6,655 establishments (2,508 + 4,147) selected with certainty.

Component ID: #ti221786952

Hypothetical Example of Certainty Payroll Cutoff

Establishment Type

Number of Establishments

Payroll ($000)

% of Total Payroll

Establishments of Multi-Establishment Firms




Single-establishment firms with payroll ≥ $229,000




Remaining “small” single-establishment firms








Component ID: #ti1877676345

The single-establishment firm sample selection had three phases: identifying the “large” single-establishment firms including some firms with special characteristics, selecting a sample of the “small” single-establishment firms, and determining if additional classification information was needed from the non-selected single-establishment firms.

Identifying “Large” Single-Establishment Firms

All single-establishment firms with annualized administrative payroll that equaled or exceeded the certainty payroll cutoff for their industry were included in the sample component of the Economic Census with certainty.  Each had a probability of selection of 1, and a sample weight of 1, which applies only for producing industry-specific statistics where data are not available from administrative records.  Note that “Large” is relative.  In some industries, this payroll cutoff was zero and all establishments were selected into the sample.

In addition, certain single-establishment firms were included with certainty, regardless of size, based on other characteristics.  These included firms that were likely cooperatives and firms included in the Annual Survey of Manufacturers (ASM).

Sampling “Small” Single-Establishment Firms

The remaining single-establishment firms (those with annualized payroll below the cutoff for their industry) were stratified by industry and state and selected using a strata-specific probability of selection.

The probabilities of selection for these strata were determined by a study of the potential respondent universe conducted shortly before sample selection operations began.  Selected small single-establishment firms were included in the sample as non-certainty cases.  Each had a probability of selection that generally fell within the range of 0.8 to 0.04.  In industry by state strata containing fewer than five establishments, all were included in the sample.

Determining Which Establishments Need Classification Information

All remaining (non-sampled) single-establishment firms with payroll were represented in the Economic Census by data from federal administrative records or through imputation and were not usually required to respond.  However, in some cases, the industry classification information on the Business Register – used to tabulate the (quantitative) administrative data in the correct industry – is inadequate or outdated.  The most common reasons for a deficient classification were administrative classification data provided to the Census Bureau lacking sufficient detail to assign an establishment to a publication level NAICS industry or the administrative data were not in agreement regarding an establishment’s classification.

Component ID: #ti1738773861

Frequency of sample redesign:

Sampling parameters (certainty payroll cutoffs, non-certainty sampling rates) are updated every five years prior to Economic Census sample selection using Business Register tabulations of payroll data and then are reviewed by subject matter experts.

Component ID: #ti24987203

Sample maintenance:

After the initial sample selection in November 2017, a second sample of single establishment companies was selected in March 2018 from those establishments on the Business Register with 2017 payroll that were not on the initial sampling frame.  Any single-establishment firms that started business so late in 2017 that their administrative data was not available to the Census Bureau in time for the last sampling operation, were included in the Census using their administrative data.

Component ID: #ti319816868

Data Collection

Component ID: #ti1416418767

Data items requested and reference period covered:

The reference period is the year 2017.  The materials and information mailed to the establishments selected into the Census are available at  The 2017 Census questionnaires are available at

Component ID: #ti122846275

Key data items:

The key statistics produced for the Economic Census include total number of establishments, value of sales, shipments, receipts, revenue; primary business activity; total number of employees; total annual payroll; total first quarter payroll; and industry specific statistics.

Component ID: #ti1623660378

Type of request:

The Economic Census is a mandatory survey and response is required by law.

Component ID: #ti1584970053

Frequency and mode of contact:

The 2017 Economic Census was the first Economic Census collected entirely online.  No paper forms were mailed to stateside businesses.  Respondents were contacted in May of 2018 with an initial survey letter requesting their participation online.  Respondents also received a reminder letter prior to the June 12 due date.  Up to four “past due” follow-ups were sent via mail and multiple email follow-ups were sent to companies that had started, but not yet completed, the reporting process online.   Select companies also received reminder calls via telephone in late 2018 and early 2019.   Initial and follow up letters can be found under respondent materials.

Component ID: #ti240027935

Data collection unit:

For all single-establishment firms and most multi-establishment firms, the data collection unit is the establishment.  As mentioned previously in the sample design section above, in certain industries, multi-establishment firms have difficulty reporting revenue and related data for each of their establishments.  These firms received a special questionnaire that requested consolidated, firm-level data for revenue and related measures.  A supplementary questionnaire listed the firm’s establishments in the industry and requested payroll and employment information for each of them.

Component ID: #ti933516864

Special procedures:

For the last non-response follow-up mailing, a certified letter was mailed to roughly one-half of the non-respondents (with the other non-respondents received a regular letter).  The non-respondents to receive the certified letter were selected using a process that identified industry by state combinations where the response was poorest and selecting a larger proportion of non-respondents from those combinations.

Component ID: #ti288083224

Compilation of Data

Component ID: #ti1899510794


Data captured in an Economic Census must be edited to identify and correct reporting errors. The data also must be adjusted to account for missing items and for businesses that do not respond. Data edits detect and validate data by considering factors such as proper classification for a given record, historical reporting for the record, and industry/geographic ratios and averages.

The first step of the data editing process is classification. To assign a valid kind-of-business or industry classification code to the establishment, computer programs subject the respondents’ responses to pre-specified items of a series of data edit programs. The specific items used for classification depend on the census report forms and include:

  • Self-designated kind-of-business check-box classifications,
  • Responses to product lines sold by a retail establishment,
  • Products manufactured by a plant, and
  • Entries written in by the respondent explaining the establishment’s activities.

If critical information is missing, the record is flagged and fixed by analysts before further processing occurs.

If all critical information is available, the classification code is assigned automatically. After classification codes are assigned, a "verification" operation is performed to validate the industry, geography and ZIP Codes.

After an establishment has been assigned a valid industry code, the data edits further evaluate the response data for consistency and validity—for example, assuring that employment data are consistent with payroll or sales/receipts data. Response data is always evaluated by industry; in some cases, type of operation or tax-exempt status is also taken into account. Additional checks compare current year data to data reported in previous censuses or from administrative sources.

Component ID: #ti1464313632


Nonresponse is defined as the inability to obtain requested data from an eligible survey unit.  Two types of nonresponse are often distinguished.  Unit nonresponse is the inability to obtain any of the substantive measurements about a unit.  In most cases of unit nonresponse, the Census Bureau was unable to obtain any information from the survey unit after several attempts to elicit a response.  Item nonresponse occurs either when a question is unanswered or the reported data is unusable.

Component ID: #ti1951790140

Nonresponse adjustment and imputation:

Nonresponse is handled by estimating or imputing missing data. Imputation is defined as the replacement of a missing or incorrectly reported item with another value derived from logical edits or statistical procedures.

The primary methods for imputing missing basic data items (such as receipts, sales, payroll, and employment) are:

  • Using administrative data
  • Deriving the missing data item from the establishment’s other data (either reported or administrative)
  • Deriving the missing data item using 2012 Census data
  • Obtaining the information from another 2017 survey

Sampled establishments that did not report NAPCS product data are assigned products in a hot-deck imputation process.  In this process the products from a similar establishment (called the donor) are assigned to the establishment missing the product data (the recipient).

Component ID: #ti1715379680

Other macro-level adjustments:

For the NAPCS product estimates, a post-stratification weighting adjustment (using industry by state strata) is subsequently performed to ensure that the weighted product data of the sampled establishments sum to the total receipts of all establishments.

Component ID: #ti1571189304

Tabulation unit:

For most statistics the unit of tabulation is the establishment.  For some tables, estimates of the number of firms is provided.

For NAPCS product tables, the sales of each individual product reported by (or imputed to) an establishment are separately tabulated.  

Component ID: #ti303613183


Economic Census tabulations for basic statistics (receipts, payroll, employment, etc.) are simple summations of data from all in-scope establishments using reported data collected from the Economic Census, plus administrative records data or imputed data for non-respondents and single-unit establishments that were not selected into the Economic Census sample.  The most common source of imputed data is administrative data from the IRS.  For multi-establishment firms in alternative reporting industries (see the sample design section above), the consolidated firm level receipts data is first allocated to the individual establishments of the firm in the industry.

Economic Census estimates for industry-specific statistics, such as NAPCS product revenue and other industry-specific special inquiries, are derived by summing weighted data, where each certainty establishment (establishments of multi-establishment firms and “large” single-establishment firms) has a weight of 1, and each non-certainty establishment has the sample weight assigned during the sample selection process (see above).  These initial weighted estimates are further adjusted to ensure that these detailed estimates sum to the basic statistics total as described above.  For example, the sales of products within an industry (estimated from the sample) should sum to the total industry receipts (computed from all establishments).

Dollar values are published in current dollars. In tables that compare the current Census to prior Census, no adjustment has been made to the estimates to account for inflation during the intervening period.

Component ID: #ti1809548351

Sampling Error:

The sampling error of an estimate based on a sample survey is the difference between the estimate and the result that would be obtained from a complete census conducted under the same survey conditions. This error occurs because characteristics differ among sampling units in the population and only a subset of the population is measured in a sample survey. The particular sample used in this survey is one of a large number of samples of the same size that could have been selected using the same sample design. Because each unit in the sampling frame had a known probability of being selected into the sample, it was possible to estimate the sampling variability of the survey estimates.

Common measures of the variability among these estimates are the sampling variance, the standard error, and the coefficient of variation (CV), which is also referred to as the relative standard error (RSE). The sampling variance is defined as the squared difference, averaged over all possible samples of the same size and design, between the estimator and its average value. The standard error is the square root of the sampling variance. The CV expresses the standard error as a percentage of the estimate to which it refers. For example, an estimate of 200 units that has an estimated standard error of 10 units has an estimated CV of 5 percent. The sampling variance, standard error, and CV of an estimate can be estimated from the selected sample because the sample was selected using probability sampling. Note that measures of sampling variability, such as the standard error and CV, are estimated from the sample and are also subject to sampling variability. It is also important to note that the standard error and CV only measure sampling variability. They do not measure any systematic biases in the estimates.

The Census Bureau recommends that individuals using these estimates incorporate sampling error information into their analyses, as this could affect the conclusions drawn from the estimates.

Estimates of basic data items, such as receipts, sales, payroll, employment, inventories, etc. included in the 2017 Economic Census First Look and Geographic Area Series publications are computed from all in-scope establishments in the country and therefore are not subject to sampling error.  For those establishments that were not sampled or did not respond, missing data items were either imputed or filled in with administrative data from other government agencies.

Component ID: #ti547231715

Nonsampling Error:

Nonsampling error encompasses all factors other than sampling error that contribute to the total error associated with an estimate. This error may also be present in censuses and other nonsurvey programs.  Nonsampling error arises from many sources, such as:

  • Inability to identify and obtain information on all units in the universe
  • Industry misclassification
  • Differences in the interpretation of the questions
  • Mismatches between sampling units and reporting units, requested data and the data available or accessible in the respondents’ records, or confusion about the period for which data is requested
  • Errors in recording, coding, or keying the data obtained, and
  • Other errors of collection, response, coverage, processing and estimation for missing or misreported data.

It is important to have metrics to measure, monitor, and manage data collection and the level of response achieved by the data collection methods so that the amount of nonresponse is minimized to the extent possible.  One type of response metric is the check-in rate.  The check-in rate is calculated as the ratio of the number of reporting units returning a questionnaire to the number of reporting units mailed a request to complete a questionnaire.  The check-in rate (expressed as a percentage) for the 2017 Economic Census was just over 75%. 

A returned questionnaire includes receipt of an electronic submission authorized by the respondent, receipt of an acceptable response during telephone follow-up, or, under special circumstances, respondent-authorized submission by some other means.

Nonsampling error for NAPCS products due to nonresponse and imputation was measured together with sampling error; coefficients of variation provided for NAPCS products include both sampling error and variability due to the imputation method used.

Although no other direct measurement of nonsampling error was obtained, precautionary steps were taken in all phases of the collection, processing, and tabulation of the data in an effort to minimize its influence.  Precise estimation of the magnitude of the nonsampling errors would require special experiments or access to independent data and, consequently, the magnitudes are often unavailable.

The Census Bureau recommends that individuals using these estimates factor in this information when assessing their analyses of these data, as nonsampling error could affect the conclusions drawn from the estimates.

Component ID: #ti491939566

Response Metrics:

For the 2017 Economic Census, the Census Bureau produced new response metrics in accordance with Census Bureau standard response rate calculations, in order to monitor data collection and to provide additional indicators of data quality.  These are the Unit Response Rate (URR), the Total Quantity Response Rate (TQRR), the Quantity Response Rate (QRR), the Administrative Data Rate (ADR), and the Imputation Rate (IR). (For definitions, see the Census Bureau Statistical Quality Standards, Appendix D3-B: Requirements for Calculating and Reporting Response Rates: Economic Surveys and Censuses, at

To produce these rates, the Census Bureau implemented a more detailed method for documenting the sources of data used for correcting estimated or inconsistent data. These correction sources align with those used in the Census Bureau’s annual economic surveys, and are defined as follows:

  1. corrections obtained directly from respondents;
  2. data from administrative sources;
  3. data from other survey(s) with the same reference period;
  4. data from another reliable secondary source for the same reference period; or
  5. estimates by industry analysts based on, for example, prior year data or industry averages.

When calculating the standard response metrics, the first four types of corrections are treated in the same manner as “reported” data.  For the fifth type of correction, the data are treated as imputed.  Census Bureau analysts’ estimates were treated as reported data for prior Economic Census reporting periods.  Data users should be advised that imputation rates calculated for the 2017 Economic Census are larger than similar metrics calculated for prior Economic Censuses, due to these procedural changes and to unit and item nonresponse in the classical sense.  The 2017 Economic Census was hampered by funding issues and delayed as a result of the partial government shutdown in January 2019.  Imputation rates are indicated with tabulated 2017 Economic Census data using a coding scheme like that presented with 2012 Economic Census data, as follows:

  • 0: Imputation rate is less than 10%
  • 1: Imputation rate is greater than or equal to 10% but less than 20%
  • 2: Imputation rate is greater than or equal to 20% but less than 30%
  • 3: Imputation rate is greater than or equal to 30% but less than 40%
  • 4: Imputation rate is greater than or equal to 40% but less than 50%
  • 5: Imputation rate is greater than or equal to 50% but less than 60%
  • 6: Imputation rate is greater than or equal to 60% but less than 70%
  • 7: Imputation rate is greater than or equal to 70% but less than 80%
  • 8: Imputation rate is greater than or equal to 80% but less than 90%
  • 9: Imputation rate is greater than or equal to 90%

Qualitative research, utilizing techniques such as cognitive interviews or usability testing methods, may be undertaken to assess the performance of new or substantially changed survey questions or data collection instruments, and results are usually used to aid design decisions in order to reduce measurement error and response burden.   Post-collection debriefing interviews may be conducted with respondents in order to evaluate the performance of questions/instruments, to identify error sources and recommend modifications for future collections.  If available, paradata may also be examined to identify problematic questions or instrument designs for further improvement.  Reports of findings and recommendations are prepared from these studies and provided to survey managers and sponsors, and may be publicly available pursuant to confidentiality and disclosure requirements.    

Component ID: #ti1045915613

Disclosure avoidance:

Disclosure is the release of data that reveals information or permits deduction of information about a particular survey unit through the release of either tables or microdata. Disclosure avoidance is the process used to protect each survey unit’s identity and data from disclosure. Using disclosure avoidance procedures, the Census Bureau modifies or removes the characteristics that put information at risk of disclosure.

Cell suppression is a disclosure avoidance technique that protects the confidentiality of individual survey units by withholding the values of certain cells within a table from release and replacing the cell value with a symbol, usually a “D”.  If the suppressed cell value were known, it would allow one to estimate an individual survey unit’s data too closely.

The cells that must be protected are called primary suppressions.  To make sure the cell values of the primary suppressions cannot be closely estimated by using other published cell values, additional cells may also be suppressed. These additional suppressed cells are called complementary suppressions.

The process of suppression does not change the higher-level totals. Values for cells that are not suppressed remain unchanged. Before the Census Bureau releases data, computer programs and analysts ensure primary and complementary suppressions have been correctly applied.

Rounding, either from the way data is collected or published, may effect whether a small number is considered a primary suppression.  Ranges are sometimes used in place of D’s to suppress sensitive data, but still provide some meaningful information.

Background on cell suppression, cell sensitivity and the protection of statistical data can be obtain from the Federal Committee on Statistical Methodology's Working Paper 22.  The Census Bureau has reviewed the 2017 Economic Census data products for unauthorized disclosure of confidential information and has approved the disclosure practices applied.  (Approval ID:  CBDRB-FY19-537)

Component ID: #ti1224664121

History of Survey Program

For more information on the history of the Economic Census, see the following page from the Census Bureau’s history site:

For more information on new content for the 2017 Economic Census and changes from the 2012 Economic Census, see the following page:

  Is this page helpful?
Thumbs Up Image Yes    Thumbs Down Image No
Comments or suggestions?
No, thanks
255 characters remaining
Thank you for your feedback.
Comments or suggestions?
Back to Header