One estimate at a time.
Each figure comes from published research, with its source, year and limits. They measure different things and are never added together.
Checked October 17, 2026. A gap shows that a difference exists; it does not by itself show why.
Explore a scenario with your own numbers.
Nothing you enter leaves your browser. Results are examples, not a record of what any family or organization has lost.
Starting values are examples. Every gap can be edited or set to zero. Domains are not added together, because their effects overlap.
Published gaps describe group averages, not a particular household. Type: user hypothetical.
Starting values are placeholders. Replace them with your own data. A difference in rates does not show why the difference exists.
The areas can overlap, so treat the total as an upper bound.
Type: organization-specific. The Opportunity Leakage Audit tests which parts of a total can be traced to practices you control.
We start with Georgia and Arkansas. State numbers are measured for the state. A national estimate is never divided by population or assigned to a state.
| Area | Source | Status |
|---|---|---|
| Work and business | BLS; Census ACS and Annual Business Survey | Not yet pulled |
| Housing and wealth | ACS; HMDA | Not yet pulled |
| Education | State education agencies; Census school finance | Not yet pulled |
| Health | CDC WONDER; state health departments | Not yet pulled |
| Mobility | Opportunity Atlas | Not yet pulled |
Six ways to work together
Rate card
| Format | In person | Virtual |
|---|---|---|
| Keynote | $3,500 + travel | $2,500 |
| Workshop | $2,000 + travel | $1,500 |
| Half-day | $4,000 + travel | $3,000 |
| Full-day | $7,000 + travel | $5,500 |
| License | $4,500 per workshop | |
| Podcast, custom work | Contact for pricing | Contact for pricing |
How we handle numbers
Every figure carries one label.
- Observed. Measured directly in a survey, census or record.
- Research estimate. Published by a study, with its method and limits.
- Modeled. Output of an economic model under stated assumptions.
- User hypothetical. A number you enter to explore a scenario.
- Organization-specific. From one client's own records.
- Causally attributable. Tied to a cause by a design that supports it.
Does every gap mean discrimination?
No. The program teaches the difference.
Is it partisan?
No. It's educational and nonpartisan.
Can it use our local data?
Yes. Sessions can include public data for your community and your own records.
Request an Institutional Assessment
We reply within 3 business days.