Employer firms
Businesses with at least one paid employee. These are the firms that create jobs, carry payroll, and show up in most economic statistics.
Nonemployer firms
Businesses with no paid employees — one person working for themselves: freelancers, contractors, gig drivers, side ventures. Most businesses in every metro are these.
Widest stable ownership gap
What this profile asks

Who is able to start businesses here? Who scales from working solo to owning a firm with employees? Who owns the firms with payroll and receipts — and where do groups show up as workers but not as owners?

Disparities appear here as facts about the region's structures and data systems — never as evidence that any group lacks ability or effort. Every chart names whose count it divides by, what it can support, what it cannot, and what further evidence a stronger claim would need. The goal is that you can read, question, and challenge every number on your own terms.

How to read this profile
Starting & sustaining Business formation, and the jump from solo work to first payroll.
Ownership & sectors Who owns employer firms, at what scale, against selectable benchmarks.
Workers vs owners The core pattern: present in the workforce, missing among owners.
Capital & markets What public data can — and mostly cannot — say about credit and contracts.
Accountability & action Mechanisms, the institutions that control them, and commitments to watch.
Methods & data quality Definitions, suppression, and everything needed to challenge us.
Business applications (state) Place Measured
About this measure
How much new-business formation intent is in the pipeline?
Counts of federal tax-ID applications, not surviving firms; no owner demographics; not seasonally adjusted. State-level context for the metro.
Nonemployer firms over time Owner Measured
About this measure
Is solo business activity growing and for whom?
Administrative tax records; gig platforms and side ventures inflate counts. Gaps in a line are years suppressed in the source data, not zeros.
Nonemployer firms per employer firm Owner Derived
About this measure
Who runs unincorporated solo activity vs firms with employees?
A high ratio can mean thriving solo entrepreneurship, blocked growth, contracting structures, or tax strategy — the ratio alone cannot say which.
Total establishments over time Place Measured
About this measure
Is the business base growing?
Establishment size structure Place Measured
About this measure
How small is the small-business base here?
Shares computed over reported classes; establishments are locations, not firms.
Firm survival by owner group Data gap

Whether firms started by owners of different groups survive at different rates in this metro is not observable in any current public dataset.

What would close this gap: Census firm-dynamics data linked to owner demographics, or a local longitudinal business registry; neither exists publicly today.

Reading this tab

Two different kinds of data live here, and telling them apart is the point. The ownership charts describe people — who owns firms, drawn from surveys that ask about the owner. The sector charts describe the place — what kinds of businesses exist here, with no information about who owns them. Public data cannot yet connect the two at the metro level: we can see that this region has, say, thousands of construction establishments, and separately that Black residents own few employer firms, but not who owns firms in which industries. That missing link is a named data gap below, not an oversight.

Ownership representation Owner Resident Worker Derived
About this measure
Who owns firms relative to who lives and works here?
Parity under any single benchmark is an analytic reference, not a natural target. Hispanic origin overlaps the race groups in every source, so shares are not additive.
Employer firms over time Owner Measured
About this measure
Is employer ownership growing and for whom?
Receipts per employer firm Owner Derived
About this measure
Do firms owned by different groups operate at similar scale?
Ownership detail, latest available years Owner Measured
Receipts in $1,000s. Blank cells with a flag are suppressed in the source data — withheld by federal disclosure rules to protect privacy, not zero.
Establishments by sector Place Measured
About this measure
What is the industry field on which local businesses operate?
Sector ownership by owner group Data gap

Which groups own firms in which sectors is suppressed at the metro level in the source survey; the sector chart above describes the place, not owners.

What would close this gap: A state-level ownership-by-industry pull joined to worker data — documented as the first extension in docs/UPDATING.md.

Worker share vs owner share Worker Owner Derived
About this measure
Where do groups participate as workers but not as owners?
Different sources, definitions, and reference periods — treat the comparison as directional. The distance between the two marks is the central pattern this application exists to show.
Average monthly earnings by worker group Worker Measured
About this measure
Are workers of different groups paid similarly in this labor market?
Employment by worker group indexed Worker Derived
About this measure
Whose employment grows or contracts through cycles?
Latest quarter worker summary Worker Measured
About this measure
Where does each group stand right now as workers?
Why this section is mostly national

For years, researchers could work with public loan-level small business lending data that carried local geography. Public access has narrowed over time, and the federal lending files that remain reliably available report national totals only — no county, no metro, no borrower demographics. Local capital access simply cannot be measured from public sources today. That changes when application-level small business lending data begin arriving under Section 1071 of the Dodd-Frank Act (first collection 2028). Until then, this tab shows the national picture in detail and names the local gaps honestly.

SBA-guaranteed lending (national) Place Measured
About this measure
What is the national volume of SBA-guaranteed lending?
Approvals only — applications and denials are not public. 7(a) dollars are total approved loan amounts; 504 dollars are the SBA/CDC-financed portion only, excluding the participating bank's share and the owner's equity. The two dollar series are therefore not comparable and are never added together here.
Recent fiscal years in detail Place Measured
National totals. Average approval is dollars divided by loan count within each program; the latest fiscal year may be partial depending on the report date.
Local lending by owner group Data gap

No application-level small business lending data by owner race exist for this region in any public file today. Community Reinvestment Act disclosures carry no borrower demographics, and the demographic fields lenders do collect are voluntary and mostly unreported.

What would close this gap: Section 1071 data: mandatory collection begins January 2028, first filings June 2029 — the first source of local application-level lending by owner group.

Procurement awards by certification status Data gap

The City of Cleveland and Cuyahoga County operate certification programs for minority- and women-owned businesses, but no current machine-readable series of contract awards by certification status was located.

What would close this gap: Annual publication of prime and subcontract awards by certification status — a proposed commitment in Accountability & action.

Interest rates, collateral, and terms Data gap

Loan pricing and collateral requirements by borrower group are unobserved in every public file covering this region.

What would close this gap: Pricing fields are not expected in the initial federal collection; voluntary lender reporting is the only nearer-term path.

Published evidence registry External evidence

External and local published evidence supporting the mechanisms examined here. Entries marked self-report describe an institution's own programs and are not independent validation of its performance. Stale entries are retained only for historical trajectory.

How to read this section

Each issue connects a regional condition to a hypothesized mechanism, the evidence for and still missing from that hypothesis, the institutions that control or influence a relevant process, and a measurable commitment that could be monitored. No institution is accused of causing a disparity; institutions are named because they control, fund, administer, or could report on a process. Every commitment below is proposed , not committed — commitments found in published documents would be labeled as such.

Full framework
Condition of the underlying data
Largest evidence gaps
  • Capital access: no local application-level lending data by owner group exist; federal collection begins 2028. See Capital & markets.
  • Procurement: no machine-readable local contract awards by certification status were located. See Accountability & action.
  • Survival and transition: firm survival and nonemployer-to-employer transition rates by owner group are not published at the metro level.
  • Sector ownership: ownership by industry is suppressed at the metro level; a state-level version is the first documented extension.
Reading this profile

Denominators differ by design. The owner surveys describe business owners . The workforce data describe workers employed in the metro. The census data describe residents . Establishment, formation, and lending files describe the place . A sector full of Black workers is not a sector full of Black-owned firms; the distance between those facts is the most useful thing this app shows.

Suppression is never zero. Cells withheld under federal disclosure rules are nulled in the pipeline and rendered as suppressed. The tally at right counts them for the selected metro; validation fails the build if a flagged cell carries a value.

Categories are constructed and imperfectly aligned. Hispanic origin overlaps race in every source. The owner survey counts White owners of any ethnicity; the resident benchmark uses White non-Hispanic. Worker race comes from linked administrative records. Cross-dataset comparisons are directional, and each panel's metadata names the specific issue.

Multi-state metros. Worker data are published as within-state metro portions; this app sums portions and employment-weights earnings, flagging merged metros in panel captions.

Units. Receipts and payroll are reported in $1,000s.

Download owner tables (CSV) Download worker table (CSV)
Suppressed cells, selected metro Owner Suppressed
Evidence-status vocabulary
Source registry and vintages
Latest validation report
Exceptions preserved during processing