Gera Civic Vitality Index (GCVI) — Methodology

Component 1 — Charities per 10,000 population (weight 45%)

For each LA: (count of active registered charities) / (ONS mid-2022 population) × 10,000. Charities are those with charity_registration_status = "Registered" in the Charity Commission extract, mapped to the LA via the geographic_area_type = "Local Authority" rows in publicextract.charity_area_of_operation.txt. A charity operating in multiple LAs is counted in each (reflects reach).

Normalisation: min-max to 0–100 across all 172 LAs, capped at the 95th-percentile value before normalisation to prevent the City of London (population ~9,000 but hundreds of national charity headquarters) from compressing the rest of the scale.

Component 2 — Income density per 10,000 population (weight 35%)

For each LA: (sum of latest_income for active charities in this LA) / (ONS mid-2022 population) × 10,000. latest_income is each charity's most recent annual return figure. Same 95th-percentile cap + min-max normalisation as Component 1.

Component 3 — Cause-mix diversity (weight 20%)

For each charity in the LA, collect its “What” cause categories from publicextract.charity_classification.txt (classification_type = “What”). The Charity Commission uses 17 cause categories (Education/training, Arts/culture, Poverty relief, etc.). Compute the Shannon entropy:

where p_i = fraction of charities in cause category i. A diversity of 1.0 would mean charities are perfectly evenly spread across all 17 categories; 0 means all charities serve a single cause. This value is computed per local authority from the real classification extract, so it varies between areas and is always below 1.0 in practice (Education/training and General Charitable Purposes are consistently over-represented). Multiply by 100 for the score.