A familiar company can appear as one business in a dataset and as several businesses in another table. The difference may come from the unit being counted: a firm or enterprise groups activity under common ownership or control, while an establishment generally represents a location where work happens. Those counts are both useful, but they answer different questions. Mixing them can make a company look like several independent businesses, or make a local industry seem smaller than its many workplaces suggest. U.S. Census Bureau U.S. Bureau of Labor Statistics

The company and the place are different units

Think of a hypothetical bakery company operating three shops in one county. The company is one firm for a firm-level count. Each shop is an establishment: a physical place where economic activity takes place. A fourth production kitchen could be another establishment even if customers never visit it. This example is only a way to explain the units, not a report about a real business. Census Bureau definitions describe a firm as one or more establishments under common ownership or control, while its Business Dynamics Statistics track activity at both levels. U.S. Census Bureau U.S. Census Bureau

A firm with one location may look like one firm and one establishment. The distinction becomes visible when a business has multiple locations, industries, or operating units. In BLS’s Quarterly Census of Employment and Wages (QCEW), an establishment is generally one economic unit at a location, often associated with one predominant kind of activity. A firm can include several such units, and each may have a different industry classification. BLS also notes that some multi-establishment employers report as a single entity, so the records are not always perfectly separated by site. U.S. Bureau of Labor Statistics

Why local counts can differ from company counts

Suppose the three bakery shops in the example are in three different towns. A location-based table can show economic activity in each town, while a firm-level table may count the parent company once. That is not a discrepancy to “fix”; the tables use different units. Establishment data are usually more suited to questions such as where workplaces are located or which industries operate in an area. Firm data can help with questions about the size or age of the organization making decisions across its locations. Census’s Business Dynamics Statistics deliberately provide both establishment-level and firm-level views. U.S. Census Bureau

Even a location label deserves a check. The Census Bureau’s nonemployer statistics concern businesses without paid employees and explain that geographic codes for many records use the owner’s mailing address, which can differ from where business activity occurs. That series treats “firm” and “establishment” interchangeably for its nonemployer data conventions. A home-based designer could therefore be assigned to a mailing-address geography, while a multi-site employer is represented through a different statistical structure. Do not assume that every dataset uses identical definitions simply because it uses the same familiar words. U.S. Census Bureau

A workplace count is not a count of owners or people

The unit also changes what a job total means. QCEW employment is reported by place of work and counts filled jobs, not unique people. A worker with two covered jobs may appear twice in the establishment totals. The measure also excludes groups outside its unemployment-insurance coverage, including self-employed workers. It is therefore useful for describing covered jobs by location and industry, but it is not a headcount of everyone who works for a company or lives in a town. U.S. Bureau of Labor Statistics

Business tables have other boundaries. Census’s Statistics of U.S. Businesses (SUSB) describes annual data for establishments with paid employees and reports firm and establishment counts, employment during the week of March 12, and annual payroll. A separate nonemployer program covers businesses without paid employees under its own criteria. Comparing an employer-only establishment count with all business tax returns, or comparing one survey’s employment measure with another, can leave out different groups and reference different periods. The table’s unit, universe, and date are part of the statistic—not fine print to skip. U.S. Census Bureau U.S. Census Bureau

A quick reading method for business statistics

Before interpreting a number of businesses, jobs, or payroll, ask four questions. First, is the row a firm, enterprise, establishment, or tax return? Second, what does the source count as a location or as common ownership? Third, which businesses and workers are in scope—employers only, a particular industry, or a broader group? Fourth, what is the reference date or period? SUSB, for example, describes employment during a particular March payroll week, while other programs may use monthly or quarterly snapshots. U.S. Census Bureau U.S. Bureau of Labor Statistics

Then match the measure to the claim. To discuss the number of workplaces in a county, use a source with establishment counts and the desired geography. To discuss the footprint of organizations across sites, use firm-level data if the source offers it. To discuss how many people have jobs, look for a person-based measure and its coverage rules rather than adding workplace counts. These checks make comparisons more honest: two official numbers can both be sound while describing different slices of economic life.

THE EVIDENCE

Sources & further reading

  1. About the Business Dynamics Statistics ↗U.S. Census Bureau
  2. Quarterly Census of Employment and Wages: Questions and Answers ↗U.S. Bureau of Labor Statistics
  3. Statistics of U.S. Businesses ↗U.S. Census Bureau
  4. Nonemployer Statistics Frequently Asked Questions ↗U.S. Census Bureau