The arguments that actually win appeals.
Analysis of 10,157 London appeals: how outcomes vary with the themes mentioned in inspector decisions, the borough and the appeal record.
This is the archive-wide baseline, not a prediction for an individual case. Among theme cohorts with at least 100 appeals, rates run from 21% to 45%; borough rates run from 17% to 43%.
Themes inferred from inspector decision text (keyword matching). Raw allow-rate = ground cited in appeals divided by those allowed.
Read: Among themes with at least 100 cases, Flood & Drainage has the highest observed allow rate (45%, n=1,109) and Housing Mix & Standards the lowest (21%, n=1,967). This is correlation, not proof that a theme caused the outcome; one decision can match several themes.
Top 8 boroughs by appeal volume. Only includes boroughs with 20+ appeals.
These are the eight largest borough cohorts. Across all qualifying boroughs, Hammersmith and Fulham is highest at 43% and Southwark lowest at 17%. The figures describe outcomes, not why they differ: case mix, appeal route and decision quality may all contribute.
A costs award can be made where a party behaved unreasonably and caused avoidable expense. This archive records whether costs were awarded, but not the assessed £ amount; the rate should not be read as evidence of motive or bad faith.
Data source: 2,666 allowed of 10,157 total London appeals (PINS appeals table, 8/11/2026, 26% allow rate).
Themes: Inferred from inspector decision text using keyword matching (e.g., "flood", "drainage", "overshadow", "heritage"). Each appeal can match multiple themes. Allow-rate = allowed appeals / total appeals mentioning that theme. Single-theme appeals are rare—most decisions weigh multiple grounds.
Boroughs: Only includes LPAs with 20+ appeals (all 33 London boroughs qualify). Sorted by volume, not by allow-rate.
Costs: the record flags appeals in which costs were awarded. An award requires unreasonable behaviour causing unnecessary expense; it is not automatic, does not by itself establish bad faith, and the assessed amount is not published in the PINS decision data.