verification / datasets / README.md

Council efficiency datasets

Working set: council pages (473 councils) built from the 100-variable shortlist selected from the full 277-variable register below.

Twelve clear-cut, easy-to-source public datasets for judging UK council efficiency and waste. All are official, per-council, and refreshed at least annually.

  1. Revenue Outturn (MHCLG) — spend by service line per council; the denominator for everything.
  2. Capital outturn + live tables on debt (MHCLG) — external debt, debt-servicing cost, asset-sale receipts.
  3. Council tax collection rates and arrears (MHCLG) — direct operational-efficiency proxy.
  4. Exceptional financial support awards + section 114 register (MHCLG / gov.uk) — official distress signals.
  5. DSG/SEND deficits (DfE, s251 returns and DSG statements) — the biggest structural deficit driver, per council.
  6. H-CLIC homelessness / temporary accommodation statistics (MHCLG) — caseload and spend on the fastest-growing cost.
  7. Adult social care finance report (DHSC, ASCFR) — unit costs in the largest service block.
  8. National Fraud Initiative (Cabinet Office) — duplicate payments, fraud and error detected per matching exercise.
  9. Housing Ombudsman + LGSCO determinations — failure demand and maladministration rates per landlord/council.
  10. Tenant Satisfaction Measures (Regulator of Social Housing) — per-landlord performance and satisfaction, annually.
  11. Transparency Code spend over £500 (each council) — supplier-level spend; supplier concentration and price dispersion.
  12. WasteDataFlow (Defra) — the cleanest unit-cost service line: cost per collection, recycling rates.

Comparator (not a ranking dataset): CIPFA nearest-neighbour groups — compare like with like (deprivation, density, rurality) or any ranking is demolished on first contact.

Methodological rule (from the leaderboard dossier): rank within one comparable kind of money at a time, normalize by council size (per resident, per pound of core spending power), and never blend one-off losses, recurring waste, and bailout stock into a single composite score.