COMMUNITY BENEFIT AGREEMENTS
There is no independent standard for what a genuinely disconnected community looks like, so there's no independent way to check whether a CBA's engagement plan actually reaches one. That's the gap this piece is built to close.
We don't verify that a Community Benefits Agreement is being honored. We supply the independent read a CBA's engagement claim can be measured against — a structural answer to "how disconnected is this community, really," produced before either party writes their own definition into the agreement.
How the Split Works
The instrument sets the disconnection standard — an independent, structural read of how trust actually moves through a specific place, produced the same way regardless of who's asking. The client's own engagement record — outreach logs, meeting attendance, survey response, whatever evidence a CBA or an underwriting process already collects — supplies the other half. The comparison between the two is the output: does the engagement record on file actually reach the community the independent read shows is most disconnected, or does it cluster somewhere easier.
For organizations building CBA frameworks and model templates: an independent disconnection standard any future agreement's engagement clause can point to, instead of leaving "meaningful" undefined.
For CDFIs and impact funds already scoring community engagement as a factor: the independent evidence layer that score currently doesn't have behind it.
Finds the census tracts most structurally disconnected from institutional trust — tested and held in the Election Day Test, a sealed, pre-registered, out-of-sample prediction.
Physical and institutional access doesn't track with income the way most engagement plans assume — confirmed across multiple cities, this shows who's structurally reachable, not who happens to answer outreach.
Election Day Test
3 of 4 sealed, pre-registered predictions held — Floor Identification confirmed out-of-sample.
Racine, Wisconsin
Neighborhood-level institutional reach ranges 16.6–58.9 — a citywide average would erase the spread entirely.
Houston, Texas
90.5% of neighborhoods carry Loyalty as the dominant value — a 28.5-point gap from what local institutions assumed and marketed on.
What this doesn't do: certify that a CBA is being complied with, adjudicate a dispute between a developer and a community, or score a specific engagement plan's quality. It reads where the disconnection actually is, at census-tract resolution, zero PII, before anyone writes an engagement clause around an assumption instead of a read.
Census tract resolution, never ZIP code. 140–145 indicators typically populated per tract, of roughly 350 in the full architecture. Refresh cadence is annual to five-year rolling averages — never real-time. Zero personally identifiable information collected at any stage.
A CBA without an independent standard is two parties negotiating over an assumption. This is the read that replaces the assumption.