In plain English: every two years the government's National Fraud Initiative cross-checks councils' records to catch payments that shouldn't have happened, and each council ends up with a pounds-found figure. This variable divides that figure by the council's size (per £1m of spending) so a small district and a big city can be compared fairly. It matters because raw fraud totals mostly reflect council size, not counter-fraud performance. Since the per-council outcomes underneath it are not published anywhere, this metric inherits that gap in full.
| name | nfi_outcomes_per_million_spend |
|---|---|
| dataset | 08-national-fraud-initiative |
| kind | prospective |
| type | numeric (ratio) |
| unit | £ outcomes per £1m net revenue expenditure |
| grain | council, per biennial NFI exercise |
| role | derived comparative metric |
| source | derived — nfi_outcomes_gbp / (service_expenditure_gbp / 1e6), with expenditure from ../01-revenue-outturn; depends on first obtaining per-council outcomes via committee-paper scrape or PSFA/audit-body request |
| period | biennial exercises (paired with matching outturn years) |
| missingness | inherits the very high missingness of nfi_outcomes_gbp |
| score_note | Size-normalised outcomes are the right comparator, but wholly dependent on nfi_outcomes_gbp. |
| peer_group | council class (district / county / unitary / metropolitan / London) — not yet scored on the leaderboard |
NFI outcomes normalised by council size, using net revenue expenditure from the revenue-outturn dataset, so large and small councils can be compared. It would indicate fraud loss identified per pound of spend. Entirely prospective — its numerator has no central per-council source — and it inherits every caveat of nfi_outcomes_gbp, including the mix of actual and estimated values and the confound with investigation effort.