Every English council must publish every payment it makes over £500 — the raw chequebook. This is simply the date each payment was made. It matters because it lets payments be grouped into months, trends be tracked over time, and near-simultaneous look-alike payments be spotted — the timeline that turns a pile of receipts into a story.
| name | payment_date |
|---|---|
| dataset | 11-transparency-code-spend |
| kind | extracted |
| type | date |
| unit | date (DD/MM/YYYY in all three sources) |
| grain | transaction |
| role | join-key |
| source | "Adur & Worthing: Posted_Date; Gloucestershire: Payment Date; Sefton: TRANSACTION DATE" |
| period | monthly publications (sample: March–June 2026 files) |
| missingness | "Derived from a three-council sample; each council's schema differs, so column mapping must be verified per council before use. All three use DD/MM/YYYY." |
| score_note | Simple, well-understood date; pure plumbing for council-month aggregation. |
| peer_group | council class (district / county / unitary / metropolitan / London) — not yet scored on the leaderboard |
The date on which the payment was posted or transacted, normalised from each council's own date column. It anchors transactions to a council-month for all constructed measures and time-based comparisons. Schema normalisation is a prerequisite: column names and header rows vary per council (Sefton's header row position differs between files), and DD/MM/YYYY parsing must not be confused with US ordering.