In plain English: every two years the government's National Fraud Initiative cross-checks councils' records to catch payments that shouldn't have happened — pensions still paid to the dead, fraudulent discounts, duplicate invoices. This number is the bottom line for one council: the pounds of fraud and error it detected or prevented as a result of that exercise. It is the most direct measure of what the cross-check actually achieved locally — and also the least-published: no central source gives it per council, so it must be dug out of committee papers or requested from the audit body.
| name | nfi_outcomes_gbp |
|---|---|
| dataset | 08-national-fraud-initiative |
| kind | prospective |
| type | numeric (GBP) |
| unit | £ per biennial exercise |
| grain | council, per biennial NFI exercise |
| role | outcome measure |
| source | not published centrally — route (b) PSFA/audit-body per-participant extract preferred; route (a) committee papers where a council publishes its own figure (outcomes in the Welsh samples appear mostly at national/Wales level, so this is the patchiest field) |
| period | biennial exercises |
| missingness | expected very high — outcome values are the least-published per-council element |
| score_note | Headline per-council money figure, but mixes actual with estimated and is barely obtainable. |
| peer_group | council class (district / county / unitary / metropolitan / London) — not yet scored on the leaderboard |
The fraud/overpayment value recorded against a council per exercise, combining detected amounts with estimated future losses prevented. It would be the headline per-council NFI result — but it does not exist in any central publication. Caveats: it mixes actual and estimated money, and because investigation is discretionary, a low value can mean little fraud or little investigation.