Spot-check agreements where amendment descriptions diverge from origin.
The dataset shows 50 federal agreements where the current-amendment description shares <30% keyword overlap with the original commitment. Drift is sometimes just rewriting (same scope, different vocabulary). Sometimes it indicates material scope change that should have triggered re-procurement. The recommendation is grounded in 50 pattern-detector matches across the federal corpus. Each cited match links back to specific source rows; an auditor walking the citation can verify every dollar in the headline.
Monetary impact
Recoverable / prevention rate (2.0%) and indirect-cost rate (1.5%) are calibrated category-level defaults grounded in published auditor benchmarks. Implementation cost is a calibrated point estimate based on category complexity. Funders should refine all three with their own historical recovery data.
Risk overview
Charity-sector regulators (CRA Charities Directorate) flag circular flows in compliance reviews; federal funders inherit reputational exposure.
Loops obscure the true cost of delivery and complicate program evaluation against intended outcomes.
TRACE-attributed methodology is now public and citable; non-action against high-score loops is itself a media risk.
Justification & audit trail
50 cited matches · $0 at stake
The dataset shows 50 federal agreements where the current-amendment description shares <30% keyword overlap with the original commitment. Drift is sometimes just rewriting (same scope, different vocabulary). Sometimes it indicates material scope change that should have triggered re-procurement. The recommendation is grounded in 50 pattern-detector matches across the federal corpus. Each cited match links back to specific source rows; an auditor walking the citation can verify every dollar in the headline.
- description_similarity0.000source: fed.grants_contributions
- description_similarity0.000source: fed.grants_contributions
- description_similarity0.000source: fed.grants_contributions
- description_similarity0.000source: fed.grants_contributions
- description_similarity0.000source: fed.grants_contributions
| Source | Row | Field | Value |
|---|---|---|---|
| fed.grants_contributions | 141-2019-2020-Q2-00850 | description_similarity | 0.000 |
| fed.grants_contributions | 141-2019-2020-Q3-00040 | description_similarity | 0.000 |
| fed.grants_contributions | 172-2018-2019-Q3-913541 | description_similarity | 0.000 |
| fed.grants_contributions | 086-2021-2022-Q2-00205 | description_similarity | 0.000 |
| fed.grants_contributions | 064-2018-2019-Q2-00056 | description_similarity | 0.000 |
Confidence rationale. Calibrated from 50 cited matches at observation severity. Confidence increases with match volume and severity; the snapshot pipeline rebuilds nightly.
Timeline & dependencies
- D+210Background monitoring established
- D+360Mid-cycle review
- D+450Decision point
Calibrated next steps
- 01Read the original and current descriptions side-by-side for each match.
- 02Where scope materially changed, flag for procurement-rule review.
- 03Cross-reference with sole-source-creep — drift + value growth is a stronger signal than drift alone.