Civometry calculation

Arrests per 100,000 residents by self-defined ethnicity

Persons arrested for notifiable offences in England and Wales (Home Office arrests table A.01c), per 100,000 usual residents in the same ethnic group (ONS Census 2021, TS021 broad groups).

How it is calculated

Numerator
Arrests for notifiable offences by self-defined ethnicity · split by self-defined ethnicity
Denominator
Census 2021 ethnic group (five broad groups) · split by Ethnic group
Formula
numerator ÷ denominator × 100,000
Result is measured in
per 100,000 residents
Which population is used
latest census on or before numerator period (Census 2021)
Suppressed when
the numerator is below 10 or the population below 1,000, so small numbers are not published as rates

How the categories are matched

The two sources do not use the same category names. Every numerator category is mapped to exactly one denominator category, and nothing is merged or dropped silently.

In the numeratorMatched to
Asian (or Asian British)Asian, Asian British or Asian Welsh
Black (or Black British)Black, Black British, Black Welsh, Caribbean or African
MixedMixed or Multiple ethnic groups
WhiteWhite
OtherOther ethnic group

Not matched: Not stated. These records have no population to divide by, so no rate is calculated for them. They are shown as counts rather than hidden.

What it covers

England and Wales as a whole (the arrests are recorded by the 43 territorial forces, but the rate is computed only at the national total); years ending March from the first period ending on or after Census day 2021. Earlier years have no census on or before them in this registry and are not computed.

44 calculated values, most recently for Year ending March 2025.

What this calculation cannot tell you

  • An arrest counts police activity, not offending; a person arrested twice is counted twice and the census counts people once.
  • The numerator's "Not stated" rows (114,657 in 2024/25, 61 per cent of Metropolitan Police records) have no denominator; every named group's rate is understated by an unknown amount and the unknown share is shown beside each value.
  • The Home Office maps arrestees to the police's 16+1 self-defined categories; the five broad groups are mapped here to the Census 2021 five broad groups by name, and people who identify as Chinese moved from Other to Asian in the police data in 2019/20.
  • The denominator is the resident population on 21 March 2021 for every year; population change after that date is not reflected.
  • Crude rates only: the age structure of each group differs and no arrests-by-age-and-ethnicity table is published, so no age-standardised rate can be calculated.

Check it yourself

Every calculated value stores the numerator, the denominator and the identifiers of the two official observations it came from. Open the numerator and denominator above, take the figures for the same period and category, and apply the formula. If you get a different answer, tell us where it diverges.