When a new Census Bureau business release lands, the headline temptation is obvious: Black-owned businesses are rising, falling, booming or lagging.
The harder question is whether the data actually support the claim.
For Black founders, lenders, investors, chambers of commerce and policymakers, the distinction matters. A count of owner-only firms does not measure job creation. A count of startup applications does not prove new Black-owned companies opened. A survey of employer firms cannot describe the full universe of side hustles, solo consultancies, barber shops, online sellers and independent contractors.
Bad denominators can turn a useful statistic into a misleading story. They can also flatten a central issue in Black business ownership: Black entrepreneurship is visible across millions of firms, but the path from self-employment to payroll, scale, contracting and generational enterprise remains uneven.
The data can tell that story. Reporters just have to use the right table.
The first split: employer firms and nonemployer firms
The most important distinction in federal business data is simple: Does the firm have paid employees?
The Census Bureau’s Annual Business Survey ↗{:target="_blank" rel="noopener"} is the primary federal source for owner demographics among employer firms, meaning businesses with paid employees. It provides estimates by race, ethnicity, sex and veteran status, along with receipts, payroll and employment.
That makes ABS the right source for questions such as:
- How many Black-owned firms have employees?
- How much revenue do Black-owned employer firms generate?
- How many workers do they employ?
- What is their annual payroll?
- Which industries have the largest Black-owned employer presence?
The Census Bureau’s Nonemployer Statistics by Demographics ↗{:target="_blank" rel="noopener"} covers a different universe: businesses with no paid employees. These are often sole proprietorships, single-member LLCs, independent professionals, gig workers, freelancers and other owner-operated firms.
That makes NES-D the better source for questions such as:
- How many Black-owned firms operate without employees?
- How much revenue do Black-owned nonemployer firms report?
- Which sectors have large concentrations of Black solo operators?
- How big is the pipeline of Black-owned firms that might one day hire?
Both categories matter. They just do not measure the same stage of business activity.
A Black woman running a home health staffing company with 12 workers belongs in the employer-firm data. A Black man operating a one-person tax preparation business belongs in the nonemployer data. Both may be legitimate businesses. Only one is an employer.
Why the headline number changes so much
The count of Black-owned firms looks very different depending on whether a story is about employer firms, nonemployer firms or both.
Employer firms are a much smaller group because hiring workers requires capital, predictable demand, compliance systems, payroll administration and management capacity. Nonemployer firms are far more numerous because many entrepreneurs begin and remain as solo operators.
That is not a weakness in the data. It is a business reality.
The problem starts when coverage uses the word “businesses” without specifying which universe it means. A story about “Black-owned businesses” that uses only employer-firm data may understate the breadth of Black entrepreneurship. A story that uses only nonemployer data may overstate the sector’s job footprint. A story that combines employer and nonemployer firms can be useful, but only if the writer says exactly what was combined and why.
For Black business coverage, this distinction is especially important because many Black-owned firms operate in sectors where owner-only or very small firms are common, including personal services, transportation, professional services, real estate, health-related services and online retail. A rise in nonemployer firms may signal opportunity, necessity, flexible work, side income or a mix of all of those. The data alone rarely tell motive.
Survey lags are not a footnote
Federal business data arrive late. That lag is normal, but it should shape every claim.
The Annual Business Survey is collected after the reference year and released after processing, estimation and review. Nonemployer data also arrive with a delay. By the time a newsroom writes about the newest available figures, the economy may have changed.
That matters during volatile periods. Pandemic-era programs, shutdowns, reopenings, inflation, interest rate increases and changes in consumer demand all affected small firms. A 2021 or 2022 business dataset may describe a specific recovery period, not the current year.
This is one reason reporters often reach for the Census Bureau’s Business Formation Statistics ↗{:target="_blank" rel="noopener"}, which track business applications more quickly. BFS data are useful for understanding startup intent and application trends. They do not identify the race of the business owner.
So a sentence like “Black entrepreneurs are filing record numbers of applications” generally cannot be supported by BFS alone unless another source links those applications to race. A more careful sentence would say business applications rose nationally, while separate lagged owner-demographic data show changes in Black-owned firms.
Race and ownership classification are more complicated than they look
Census business ownership data usually describe firms by majority ownership. In practical terms, a business is generally classified as Black or African American-owned when Black or African American owners hold at least 51% of the equity, interest or stock, subject to the Census Bureau’s survey definitions and classification rules.
That matters for partnerships.
A firm with a Black founder who owns 50% and a non-Black co-founder who owns 50% may not appear as majority Black-owned, even if the Black founder is the CEO and public face of the company. A publicly held company led by a Black chief executive is not the same thing as a Black-owned firm. A franchise operated by a Black owner may count if the local business meets ownership criteria, while the parent brand may not.
Race and ethnicity are also separate concepts in federal statistics. The Census Bureau follows federal standards for race and ethnicity, which have been governed by the Office of Management and Budget’s Statistical Policy Directive No. 15 ↗{:target="_blank" rel="noopener"}. A person can be Black and Hispanic. A business table may present race, ethnicity or combinations depending on the product and year.
Good coverage should use the label in the source. If the table says “Black or African American-owned,” use that language. If the data refer to “minority-owned,” do not automatically treat that as a Black business statistic. If the source counts “classifiable firms,” explain that some firms cannot be assigned to an owner-demographic category.
What each metric can and cannot prove
A strong Black business data story usually needs more than a firm count. The right metrics depend on the claim.
Firm count
Use firm count to describe presence, access and entrepreneurial activity. It answers: How many businesses exist in this category?
But firm count alone does not measure scale. One firm may be a solo consultant with $40,000 in annual receipts. Another may employ 80 people and generate $20 million in revenue. Counting both as one business is statistically valid, but it does not describe economic weight.
Receipts
Receipts show reported sales or revenue. They help measure market activity and scale.
For Black-owned firms, receipts can reveal a gap that firm counts hide. A group may have a growing number of businesses but a much smaller share of receipts, which points to differences in firm size, capital access, industry concentration and customer base.
When possible, compare receipts within the same dataset, year, geography and industry. Avoid comparing average receipts across broad sectors without noting that industries have very different cost structures.
Employment and payroll
Use employment and payroll only for employer firms. These metrics answer job creation and wage-footprint questions.
If an article says Black-owned businesses employ a certain number of workers, the source should be employer-firm data, not nonemployer data. Nonemployer firms, by definition, do not have paid employees.
Average receipts per firm
This metric can be helpful, but it requires caution. An average can be pulled upward by a small number of larger firms. If median data are available, they often provide a better picture of the typical firm. If only averages are available, do not describe them as typical without qualification.
Sector concentration
Industry data can show where Black-owned firms are clustered. That is useful for understanding exposure to economic shocks and opportunities. A downturn in transportation, home care, beauty services or food service will not affect all owner groups equally if their industry mix differs.
Geography
State and metro data can be powerful, but smaller samples often mean wider uncertainty or suppressed estimates. A local ranking may look precise while resting on limited data. Reporters should check margins of error, suppression notes and whether the source supports the level of geography being used.
Common mistakes in Black business data stories
Several errors show up repeatedly in coverage of Black-owned firms.
One is treating business applications as operating businesses. An application may become a business, but it may also never launch, remain inactive or represent a tax or legal restructuring.
Another is treating nonemployer growth as proof of broad business scaling. More owner-only firms may be encouraging, but it does not automatically mean more firms are hiring.
A third is comparing datasets that use different definitions. The Annual Business Survey, Nonemployer Statistics by Demographics, Business Formation Statistics and the Census Bureau’s Statistics of U.S. Businesses ↗{:target="_blank" rel="noopener"} each answer different questions. SUSB is strong for employer firm size and industry dynamics, but it does not provide owner race. ABS provides owner demographics for employer firms. NES-D provides owner demographics for nonemployers. BFS tracks applications.
A fourth mistake is ignoring uncertainty. ABS is a survey, so estimates may have sampling error. The Census Bureau publishes technical documentation for the Annual Business Survey methodology ↗{:target="_blank" rel="noopener"} and Nonemployer Statistics methodology ↗{:target="_blank" rel="noopener"}. When changes are small, the question is not only whether the number moved. It is whether the movement is meaningful.
The Black business story is not one number
The most accurate Black business stories usually hold two ideas at once.
First, Black entrepreneurship is broad and visible, especially when nonemployer firms are included. Millions of Black owners operate businesses that generate income, serve communities and build experience.
Second, the employer-firm universe remains the critical measure for scale. Employer firms drive payroll, hiring, contracting capacity and larger revenue footprints. The transition from solo operation to employer business is a major threshold, and federal data show why it deserves attention.
That threshold is not just a statistical line. It reflects access to credit, reliable customers, management capacity, procurement opportunities, insurance, real estate, technology, networks and time. The Census tables will not explain every barrier. They can show where the gaps appear.
A better newsroom standard
When writing about Black-owned business growth, every story should answer five questions before publication:
1. Is this employer-firm data, nonemployer data, application data or a combination? 2. What year does the data describe? 3. How does the source define Black ownership? 4. Are the comparisons from the same dataset and geography? 5. Does the claim match the metric?
If the story is about jobs, use employer firms, employees and payroll. If it is about entrepreneurial participation, include nonemployers. If it is about startup momentum, be clear when using business applications and do not assign race where the data do not.
Black business coverage does not need smaller claims. It needs sharper ones.