National Fraud Initiative (NFI) — Cabinet Office / Public Sector Fraud Authority

Publisher and landing page

• Publisher: Public Sector Fraud Authority (Cabinet Office / HM Treasury). Earlier editions: Cabinet Office; pre-2015: Audit Commission.
• Official landing page: https://www.gov.uk/government/publications/national-fraud-initiative-reports (corporate report collection, last updated 2025-03-12)
• Wider collection: https://www.gov.uk/government/collections/national-fraud-initiative

Key finding: NO per-council bulk data file exists

The Cabinet Office publishes NFI outcomes only as national-level PDF reports (England totals, broken down by risk area and by nation — Scotland, Wales and NI run parallel exercises via their audit bodies). The full attachment list of the reports publication (checked via the gov.uk content API on 2026-08-13) contains only PDFs and HTML report pages back to 2000 — no CSV/ODS, and no per-participant table inside the PDFs (Appendices 1–3 of the 2022–24 report are England-wide by case type / risk area).

Per-council match outcomes are returned to each participating body through the secure NFI web application and are not published centrally. Individual councils typically report their own NFI results to their audit/governance committees. A full per-council collection would require scraping ~300+ councils' committee-paper repositories (ModernGov etc.) for their biennial NFI outcome reports, or an FOI/data request to the PSFA — neither yields a consistent machine-readable dataset.

Files downloaded

All in data/ (downloaded 2026-08-13):

Representative per-council samples (councils' own audit-committee NFI reports), in data/samples/:

Granularity, coverage, period

• Granularity: national report per biennial matching exercise; England outcomes by risk area (housing benefit, council tax SPD/reduction, pensions, payroll, creditor/duplicate payments, blue badges, etc.). NOT per council.
• Geographic coverage: UK-wide programme; the Cabinet Office report focuses on England (£477.5m of the £510m 2022–24 UK total). Scotland: Audit Scotland; Wales: Audit Wales; NI: NI Audit Office publish their own NFI reports.
• Time period: latest report covers April 2022 – March 2024; back editions to 1996/98 (first report May 2000).

Refresh cadence

Biennial. The main data-matching exercise runs every two years; the national report follows roughly a year after the exercise closes. Next report (2024–2026 exercise) expected around early 2027 (unconfirmed — inferred from the 2022→2025 pattern).

Re-download

curl -L -o NFI_report_2022-24.pdf \
  https://assets.publishing.service.gov.uk/media/67b6f6f89ae06ef4a71cf2ff/2025-01-28_NFI_report_2022-24_v3.pdf
curl -L -o NFI_report_2020-22_dec2022.pdf \
  https://assets.publishing.service.gov.uk/media/638a131d8fa8f569f6a9d252/2022-12-02_NFI_report_2022__12v3__-_JQ.pdf

Attachment list can be re-derived from: https://www.gov.uk/api/content/government/publications/national-fraud-initiative-reports

Licence

Open Government Licence v3.0 — confirmed: the gov.uk publication page states "Open Government Licence" and the 2022–24 report PDF text includes the OGL statement.

Caveats

• Headline figures mix detected fraud/error with estimated future losses prevented (Appendix 3 of the 2022–24 report explains the estimation basis) — do not treat as pure recovered cash.
• 2022–24 total (£510m UK) includes £139.5m from the private-sector "Business Development" workstream, not comparable with earlier editions.
• Machine-of-record for per-council figures is the secure NFI app; council committee reports (like the samples) vary in format, period and completeness, and Welsh/Scottish councils are reported by their own audit bodies.
• Cannot be used directly for a per-council efficiency ranking without a bespoke collection exercise (committee-paper scraping or PSFA data request).