Skip to content

How do arrests, convictions and prison numbers in England and Wales differ by ethnicity?

In 2025/26 the Home Office recorded 10.57 arrests per 1,000 people who define themselves as White, 22.87 Black, 10.73 Asian, 14.76 Mixed and 11.44 Other (ethnicity not stated for 11.13% of arrests); of people convicted of indictable offences in 2025 whose ethnicity was recorded (28.39% unknown), 82.1% were White, 7.29% Black, 5.59% Asian, 3.51% Mixed and 1.51% Other, against 82.82%, 3.91%, 8.91%, 2.32% and 2.04% of residents aged 10 and over in the 2021 census; and on 30 June 2026, 72.01% of prisoners with a stated ethnicity were White, 12.09% Black, 8.47% Asian, 4.99% Mixed and 2.45% Other.

Arrests per 1,000 people: White (self-defined)

10.57

arrests per 1,000 people who define themselves as White, Year ending 31 March 2026

Moderate trust, 63 of 100 Only Home Office counts: EF&F 2006-20 republishes it, so there is nothing independent to check it against.

Each source by year: arrests per 1,000 people who define themselves as White
  • Home Office
  • EF&F 2006-20
0102030200720102015202020262020: Home Office: From 2019/20, arrests of people who identify as Chinese are counted under Asian (or Asian British); before that they were in Other (Home Office note 3). · EF&F 2006-20: For 2019/20 the file uses the 2011 census grouping (Chinese in Asian); earlier years use the 2001 grouping (Chinese in Other).2021: Home Office: Rates per 1,000 from 2020/21 use 2021 census populations; earlier rates used 2011 census populations and are not comparable.2023: Home Office: In November 2022 the Metropolitan Police changed data system; the share of its arrest records without self-defined ethnicity rose from 4% in 2021/22 to 69% in 2023/24 and was 34% in 2025/26 (Home Office note 14). · Home Office: Arrests by officer-observed ethnicity published from 2022/23, alongside self-defined ethnicity (Table A_01d).2025: Home Office: Derbyshire Constabulary could not provide data for 2024/25 and 2025/26 because of data quality issues; its data are excluded from time series comparisons (Home Office note 10).2018: EF&F 2006-20: Lancashire could not submit arrests data for some years in the 2015 to 2020 file; the page adds an "All - excluding Lancashire" area for comparison.

Recorded change in how a source reports; hover or focus the ring for details.

Show the data as a table
YearHome Office, Police powers and procedures: arrests ethnic disparity time series, year ending March 2026Ethnicity facts and figures (Cabinet Office), Arrests: data files 2006 to 2020
2007–23.51
2008–24.28
2009–24.45
2010–22.88
2011–22.31
2012–20.18
2013–17.53
2014–16.59
2015–15.15
2016–13.46
2017–11.46
2018–10.36
2019–9.6
2020–9.1
20218.63–
20228.88–
20239.39–
20249.82–
202510.08–
202610.57–
What this number counts
Counts
Arrests for notifiable offences per 1,000 residents of the White group (self-defined ethnicity)
Where
England and Wales, excluding Derbyshire
When
Year ending 31 March 2026; stored under the year the financial year ends (one year, not an average)

The figure from Home Office, the only source that counts here.

Where the number comes from (2 sources)
SourceFigure and yearDefinitionQualityCounts for
Home Office, Police powers and procedures: arrests ethnic disparity time series, year ending March 2026Home Office · Police and justice recordsRetrieved 2026-10-27 · Data (spreadsheet)
10.572026
Arrests for notifiable offences of people who defined their ethnicity as White, per 1,000 residents of that group in the 2021 census, England and Wales excluding Derbyshire, year ending 31 March 2026: 504,682 arrests and a population of 47,741,067.

A person is counted each time they are arrested for an unconnected offence in the year. Arrests with ethnicity not stated are in no group.

Same value in summary Table A.04 (https://assets.publishing.service.gov.uk/media/6abbcced3fe1a15db6aff295/arrests-data-tables-summary-mar26.ods). 2025/26 share of arrests with a stated ethnicity: 76.77%. Publication: https://www.gov.uk/government/statistics/stop-and-search-arrests-and-mental-health-detentions-march-2026.

4 of 5 checks
  • Yes: Method and definitions published. Each release has notes on what counts as an arrest, the self-defined ethnic groups and the census population used.
  • Yes: Official statistics or peer-reviewed. Official statistics published by the Home Office (Police powers and procedures).
  • Yes: Full count or representative sample. Every arrest for a notifiable offence reported by the police forces; Derbyshire is left out of the series because it could not supply 2024/25 and 2025/26 data.
  • Yes: Raw data downloadable. Counts, populations and rates are in open ODS tables.
  • No: No known break or quality problem. Ethnicity was not stated for 11.13% of arrests in 2025/26, varying between 1.88% and 17.4% since 2006/07. In November 2022 the Metropolitan Police changed data system; the share of its arrest records without self-defined ethnicity rose from 4% in 2021/22 to 69% in 2023/24 and was 34% in 2025/26 (Home Office note 14).
100%
Ethnicity facts and figures (Cabinet Office), Arrests: data files 2006 to 2020Cabinet Office Race Disparity Unit (Home Office data) · Police and justice recordsRetrieved 2026-10-27
9.12020
Home Office arrest counts for the White group (self-defined) divided by 2011 census populations, England and Wales, years ending March 2007 to 2020.

Census 2011 population used: 48,209,395. Not comparable with the 2021-census rates from 2020/21.

2017/18 and 2018/19 exclude Lancashire (counts and population).

4 of 5 checks
  • Yes: Method and definitions published. The page explains the counting rules, the ethnic groups and the census base for the rates.
  • Yes: Official statistics or peer-reviewed. Home Office official statistics, republished with rates by the Cabinet Office Race Disparity Unit.
  • Yes: Full count or representative sample. Every arrest for a notifiable offence reported by the police forces (City of London excluded).
  • Yes: Raw data downloadable. Downloadable CSV files with counts, populations and rates.
  • No: No known break or quality problem. Every year uses 2011 census populations; Lancashire could not supply data for some years, so 2017/18 and 2018/19 exclude it; Chinese people moved from Other to Asian in 2019/20.
Not counted

Better and more direct sources count for more. A source that only copies another does not count.

  • Home Office100%

Shown but not counted: EF&F 2006-20 (republishes Home Office).

How the trust score was worked out
Quality32 / 40

How well the sources are made

Agreement15 / 30

Unknown with one source, so set to half

Independence6 / 20

How many separate ways the number was measured

Freshness10 / 10

How recent the newest data is

How the score works

Who reports it, and since when

Which kinds of source measure this number, which years each one published, and where the reporting stops or has holes.

No source of these kinds: victim survey, independent count.

  • Single source: nothing to check it against
  • EF&F 2006-20: no figure here after 2020, while others reach 2026
  • Home Office, 2020: From 2019/20, arrests of people who identify as Chinese are counted under Asian (or Asian British); before that they were in Other (Home Office note 3). source
  • Home Office, 2021: Rates per 1,000 from 2020/21 use 2021 census populations; earlier rates used 2011 census populations and are not comparable. source
  • Home Office, 2023: In November 2022 the Metropolitan Police changed data system; the share of its arrest records without self-defined ethnicity rose from 4% in 2021/22 to 69% in 2023/24 and was 34% in 2025/26 (Home Office note 14). source
  • Home Office, 2023: Arrests by officer-observed ethnicity published from 2022/23, alongside self-defined ethnicity (Table A_01d). source
  • Home Office, 2025: Derbyshire Constabulary could not provide data for 2024/25 and 2025/26 because of data quality issues; its data are excluded from time series comparisons (Home Office note 10). source
  • EF&F 2006-20, 2018: Lancashire could not submit arrests data for some years in the 2015 to 2020 file; the page adds an "All - excluding Lancashire" area for comparison. source
  • EF&F 2006-20, 2020: For 2019/20 the file uses the 2011 census grouping (Chinese in Asian); earlier years use the 2001 grouping (Chinese in Other). source
Years each source has a figure here
'07'10'15'20'25'26Home OfficeEF&F 2006-20

figure no figure here, inside its range no later figure recorded change

Arrests per 1,000 people: Black (self-defined)

22.87

arrests per 1,000 people who define themselves as Black or Black British, Year ending 31 March 2026

Moderate trust, 63 of 100 Only Home Office counts: EF&F 2006-20 republishes it, so there is nothing independent to check it against.

Each source by year: arrests per 1,000 people who define themselves as Black or Black British
  • Home Office
  • EF&F 2006-20
020406080200720102015202020262020: Home Office: From 2019/20, arrests of people who identify as Chinese are counted under Asian (or Asian British); before that they were in Other (Home Office note 3). · EF&F 2006-20: For 2019/20 the file uses the 2011 census grouping (Chinese in Asian); earlier years use the 2001 grouping (Chinese in Other).2021: Home Office: Rates per 1,000 from 2020/21 use 2021 census populations; earlier rates used 2011 census populations and are not comparable.2023: Home Office: In November 2022 the Metropolitan Police changed data system; the share of its arrest records without self-defined ethnicity rose from 4% in 2021/22 to 69% in 2023/24 and was 34% in 2025/26 (Home Office note 14). · Home Office: Arrests by officer-observed ethnicity published from 2022/23, alongside self-defined ethnicity (Table A_01d).2025: Home Office: Derbyshire Constabulary could not provide data for 2024/25 and 2025/26 because of data quality issues; its data are excluded from time series comparisons (Home Office note 10).2018: EF&F 2006-20: Lancashire could not submit arrests data for some years in the 2015 to 2020 file; the page adds an "All - excluding Lancashire" area for comparison.Home OfficeEF&F 2006-20

Recorded change in how a source reports; hover or focus the ring for details.

Show the data as a table
YearHome Office, Police powers and procedures: arrests ethnic disparity time series, year ending March 2026Ethnicity facts and figures (Cabinet Office), Arrests: data files 2006 to 2020
2007–56.38
2008–58.56
2009–59.67
2010–59.51
2011–60.64
2012–54.87
2013–47.48
2014–46.25
2015–44.69
2016–43.91
2017–38.04
2018–34.78
2019–32.3
2020–31.4
202122.62–
202221.17–
202320.37–
202416.41–
202518.56–
202622.87–
What this number counts
Counts
Arrests for notifiable offences per 1,000 residents of the Black group (self-defined ethnicity)
Where
England and Wales, excluding Derbyshire
When
Year ending 31 March 2026; stored under the year the financial year ends (one year, not an average)

The figure from Home Office, the only source that counts here.

Where the number comes from (2 sources)
SourceFigure and yearDefinitionQualityCounts for
Home Office, Police powers and procedures: arrests ethnic disparity time series, year ending March 2026Home Office · Police and justice recordsRetrieved 2026-10-27 · Data (spreadsheet)
22.872026
Arrests for notifiable offences of people who defined their ethnicity as Black or Black British, per 1,000 residents of that group in the 2021 census, England and Wales excluding Derbyshire, year ending 31 March 2026: 54,779 arrests and a population of 2,394,720.

A person is counted each time they are arrested for an unconnected offence in the year. Arrests with ethnicity not stated are in no group.

Same value in summary Table A.04 (https://assets.publishing.service.gov.uk/media/6abbcced3fe1a15db6aff295/arrests-data-tables-summary-mar26.ods). 2025/26 share of arrests with a stated ethnicity: 8.33%. Publication: https://www.gov.uk/government/statistics/stop-and-search-arrests-and-mental-health-detentions-march-2026.

4 of 5 checks
  • Yes: Method and definitions published. Each release has notes on what counts as an arrest, the self-defined ethnic groups and the census population used.
  • Yes: Official statistics or peer-reviewed. Official statistics published by the Home Office (Police powers and procedures).
  • Yes: Full count or representative sample. Every arrest for a notifiable offence reported by the police forces; Derbyshire is left out of the series because it could not supply 2024/25 and 2025/26 data.
  • Yes: Raw data downloadable. Counts, populations and rates are in open ODS tables.
  • No: No known break or quality problem. Ethnicity was not stated for 11.13% of arrests in 2025/26, varying between 1.88% and 17.4% since 2006/07. In November 2022 the Metropolitan Police changed data system; the share of its arrest records without self-defined ethnicity rose from 4% in 2021/22 to 69% in 2023/24 and was 34% in 2025/26 (Home Office note 14).
100%
Ethnicity facts and figures (Cabinet Office), Arrests: data files 2006 to 2020Cabinet Office Race Disparity Unit (Home Office data) · Police and justice recordsRetrieved 2026-10-27
31.42020
Home Office arrest counts for the Black group (self-defined) divided by 2011 census populations, England and Wales, years ending March 2007 to 2020.

Census 2011 population used: 1,864,890. Not comparable with the 2021-census rates from 2020/21.

2017/18 and 2018/19 exclude Lancashire (counts and population).

4 of 5 checks
  • Yes: Method and definitions published. The page explains the counting rules, the ethnic groups and the census base for the rates.
  • Yes: Official statistics or peer-reviewed. Home Office official statistics, republished with rates by the Cabinet Office Race Disparity Unit.
  • Yes: Full count or representative sample. Every arrest for a notifiable offence reported by the police forces (City of London excluded).
  • Yes: Raw data downloadable. Downloadable CSV files with counts, populations and rates.
  • No: No known break or quality problem. Every year uses 2011 census populations; Lancashire could not supply data for some years, so 2017/18 and 2018/19 exclude it; Chinese people moved from Other to Asian in 2019/20.
Not counted

Better and more direct sources count for more. A source that only copies another does not count.

  • Home Office100%

Shown but not counted: EF&F 2006-20 (republishes Home Office).

How the trust score was worked out
Quality32 / 40

How well the sources are made

Agreement15 / 30

Unknown with one source, so set to half

Independence6 / 20

How many separate ways the number was measured

Freshness10 / 10

How recent the newest data is

How the score works

Who reports it, and since when

Which kinds of source measure this number, which years each one published, and where the reporting stops or has holes.

No source of these kinds: victim survey, independent count.

  • Single source: nothing to check it against
  • EF&F 2006-20: no figure here after 2020, while others reach 2026
  • Home Office, 2020: From 2019/20, arrests of people who identify as Chinese are counted under Asian (or Asian British); before that they were in Other (Home Office note 3). source
  • Home Office, 2021: Rates per 1,000 from 2020/21 use 2021 census populations; earlier rates used 2011 census populations and are not comparable. source
  • Home Office, 2023: In November 2022 the Metropolitan Police changed data system; the share of its arrest records without self-defined ethnicity rose from 4% in 2021/22 to 69% in 2023/24 and was 34% in 2025/26 (Home Office note 14). source
  • Home Office, 2023: Arrests by officer-observed ethnicity published from 2022/23, alongside self-defined ethnicity (Table A_01d). source
  • Home Office, 2025: Derbyshire Constabulary could not provide data for 2024/25 and 2025/26 because of data quality issues; its data are excluded from time series comparisons (Home Office note 10). source
  • EF&F 2006-20, 2018: Lancashire could not submit arrests data for some years in the 2015 to 2020 file; the page adds an "All - excluding Lancashire" area for comparison. source
  • EF&F 2006-20, 2020: For 2019/20 the file uses the 2011 census grouping (Chinese in Asian); earlier years use the 2001 grouping (Chinese in Other). source
Years each source has a figure here
'07'10'15'20'25'26Home OfficeEF&F 2006-20

figure no figure here, inside its range no later figure recorded change

Arrests with self-defined ethnicity not stated

11.13

% of arrests with self-defined ethnicity not stated, Year ending 31 March 2026

Moderate trust, 63 of 100 Only one source (Home Office), so it cannot be checked against another.

Each source by year: % of arrests with self-defined ethnicity not stated
05101520200720102015202020262020: Home Office: From 2019/20, arrests of people who identify as Chinese are counted under Asian (or Asian British); before that they were in Other (Home Office note 3).2021: Home Office: Rates per 1,000 from 2020/21 use 2021 census populations; earlier rates used 2011 census populations and are not comparable.2023: Home Office: In November 2022 the Metropolitan Police changed data system; the share of its arrest records without self-defined ethnicity rose from 4% in 2021/22 to 69% in 2023/24 and was 34% in 2025/26 (Home Office note 14). · Home Office: Arrests by officer-observed ethnicity published from 2022/23, alongside self-defined ethnicity (Table A_01d).2025: Home Office: Derbyshire Constabulary could not provide data for 2024/25 and 2025/26 because of data quality issues; its data are excluded from time series comparisons (Home Office note 10).

Recorded change in how a source reports; hover or focus the ring for details.

Show the data as a table
YearHome Office, Police powers and procedures: arrest summary tables, year ending March 2026, Table A_01c
20077.31
20084.06
20092.27
20102.39
20111.89
20121.88
20132.09
20142.5
20152.58
20165.41
20176.72
20186.71
201911.83
202015.13
202115.02
202214.53
202312.65
202417.4
202515.99
202611.13
What this number counts
Counts
Share of arrests for notifiable offences where the person's self-defined ethnicity is not stated
Where
England and Wales, excluding Derbyshire
When
Year ending 31 March 2026 (one year, not an average)

The figure from Home Office, the only source that counts here.

Where the number comes from (1 source)
SourceFigure and yearDefinitionQualityCounts for
Home Office, Police powers and procedures: arrest summary tables, year ending March 2026, Table A_01cHome Office · Police and justice recordsRetrieved 2026-10-27 · Data (spreadsheet)
11.132026
"Not stated" arrests divided by all arrests: 82,321 of 739,703 in 2025/26.
4 of 5 checks
  • Yes: Method and definitions published. Each release has notes on what counts as an arrest, the self-defined ethnic groups and the census population used.
  • Yes: Official statistics or peer-reviewed. Official statistics published by the Home Office (Police powers and procedures).
  • Yes: Full count or representative sample. Every arrest for a notifiable offence reported by the police forces; Derbyshire is left out of the series because it could not supply 2024/25 and 2025/26 data.
  • Yes: Raw data downloadable. Counts, populations and rates are in open ODS tables.
  • No: No known break or quality problem. Ethnicity was not stated for 11.13% of arrests in 2025/26, varying between 1.88% and 17.4% since 2006/07. In November 2022 the Metropolitan Police changed data system; the share of its arrest records without self-defined ethnicity rose from 4% in 2021/22 to 69% in 2023/24 and was 34% in 2025/26 (Home Office note 14).
100%

Better and more direct sources count for more. A source that only copies another does not count.

  • Home Office100%
How the trust score was worked out
Quality32 / 40

How well the sources are made

Agreement15 / 30

Unknown with one source, so set to half

Independence6 / 20

How many separate ways the number was measured

Freshness10 / 10

How recent the newest data is

How the score works

Who reports it, and since when

Which kinds of source measure this number, which years each one published, and where the reporting stops or has holes.

No source of these kinds: victim survey, independent count.

  • Single source: nothing to check it against
  • Home Office, 2020: From 2019/20, arrests of people who identify as Chinese are counted under Asian (or Asian British); before that they were in Other (Home Office note 3). source
  • Home Office, 2021: Rates per 1,000 from 2020/21 use 2021 census populations; earlier rates used 2011 census populations and are not comparable. source
  • Home Office, 2023: In November 2022 the Metropolitan Police changed data system; the share of its arrest records without self-defined ethnicity rose from 4% in 2021/22 to 69% in 2023/24 and was 34% in 2025/26 (Home Office note 14). source
  • Home Office, 2023: Arrests by officer-observed ethnicity published from 2022/23, alongside self-defined ethnicity (Table A_01d). source
  • Home Office, 2025: Derbyshire Constabulary could not provide data for 2024/25 and 2025/26 because of data quality issues; its data are excluded from time series comparisons (Home Office note 10). source
Years each source has a figure here
'07'10'15'20'25'26Home Office

figure no figure here, inside its range no later figure recorded change

Arrests per 1,000 people, by self-defined ethnicity (years ending March 2007-2026)

Each line is the source that counts most for that group.

  • White
  • Black or Black British
  • Asian or Asian British
  • Mixed
  • Other Ethnic group
0102030202120222023202420252026
GroupFigureYearTrust
White10.57202663 Moderate
Black or Black British22.87202663 Moderate
Asian or Asian British10.73202663 Moderate
Mixed14.76202663 Moderate
Other Ethnic group11.44202663 Moderate
Show every year as a table
YearWhiteBlack or Black BritishAsian or Asian BritishMixedOther Ethnic group
20218.6322.628.1311.857.84
20228.8821.177.611.567.73
20239.3920.378.2712.448.44
20249.8216.418.1511.948.28
202510.0818.569.0412.629.62
202610.5722.8710.7314.7611.44

Share of people convicted of indictable offences, % with a recorded ethnicity (2010-2025)

Each line is the source that counts most for that group.

  • White
  • Black
  • Asian
  • Mixed
  • Other
02550751002010201520202025
GroupFigureYearTrust
White82.1202563 Moderate
Black7.29202563 Moderate
Asian5.59202563 Moderate
Mixed3.51202563 Moderate
Other1.51202563 Moderate
Show every year as a table
YearWhiteBlackAsianMixedOther
201082.158.415.33.071.07
201181.858.615.293.161.1
201282.158.415.253.11.09
201382.048.465.263.081.17
201482.268.495.212.981.06
201581.48.845.453.171.15
201679.6510.086.043.091.13
201779.4510.326.063.031.14
201879.8610.135.73.171.13
201978.7510.486.043.431.31
202078.9410.225.893.571.38
202178.3210.46.253.541.5
202278.769.886.213.631.52
202382.417.235.493.511.37
202483.376.65.283.341.41
202582.17.295.593.511.51

Share of people convicted of sexual offences, % with a recorded ethnicity (2010-2025)

Each line is the source that counts most for that group.

  • White
  • Black
  • Asian
  • Mixed
  • Other
02550751002010201520202025
GroupFigureYearTrust
White81.74202563 Moderate
Black5.94202563 Moderate
Asian8.36202563 Moderate
Mixed2.34202563 Moderate
Other1.62202563 Moderate
Show every year as a table
YearWhiteBlackAsianMixedOther
201081.287.238.561.431.5
201181.127.098.821.641.33
201279.747.959.231.621.47
201379.647.219.481.662.01
201480.656.859.481.721.3
201580.157.119.371.731.64
201680.946.678.991.671.74
201780.668.067.821.741.72
201881.767.48.211.281.35
201983.586.637.091.271.43
202084.396.536.241.751.1
202183.496.346.951.661.57
202282.26.547.712.231.33
202381.836.187.892.072.02
202482.525.818.421.751.49
202581.745.948.362.341.62

Share of the prison population, % with a stated ethnicity (30 June 2005-2026)

Each line is the source that counts most for that group.

  • White
  • Black or Black British
  • Asian or Asian British
  • Mixed
  • Other ethnic group
02040608020052010201520202026
GroupFigureYearTrust
White72.01202663 Moderate
Black or Black British12.09202663 Moderate
Asian or Asian British8.47202663 Moderate
Mixed4.99202663 Moderate
Other ethnic group2.45202663 Moderate
Show every year as a table
YearWhiteBlack or Black BritishAsian or Asian BritishMixedOther ethnic group
200575.1914.786.042.851.14
200673.915.276.622.971.23
200773.5915.186.683.081.46
200872.8715.27.033.261.64
200973.0214.557.263.421.74
201073.67147.273.61.45
201174.3513.367.443.741.11
20127413.497.583.781.15
201373.8113.197.873.91.23
201474.0212.787.844.121.25
201574.212.498.174.081.06
201674.0112.368.114.331.18
201773.612.58.24.461.24
201873.2612.658.064.581.44
201972.8412.858.074.691.55
202072.6912.967.994.871.5
202172.4212.868.34.931.48
202272.6212.68.184.961.64
202372.7412.358.054.891.98
202472.4812.198.294.832.21
202572.6211.958.034.872.53
202672.0112.098.474.992.45

Share of residents aged 10 and over, by ethnic group (census 2011 and 2021)

Each line is the source that counts most for that group.

  • White
  • Black, Black British, Black Welsh, Caribbean or African
  • Asian, Asian British or Asian Welsh
  • Mixed or Multiple ethnic groups
  • Other ethnic group
02550751002011201220142016201820202021
GroupFigureYearTrust
White82.82202168 Moderate
Black, Black British, Black Welsh, Caribbean or African3.91202168 Moderate
Asian, Asian British or Asian Welsh8.91202168 Moderate
Mixed or Multiple ethnic groups2.32202168 Moderate
Other ethnic group2.04202168 Moderate
Show every year as a table
YearWhiteBlack, Black British, Black Welsh, Caribbean or AfricanAsian, Asian British or Asian WelshMixed or Multiple ethnic groupsOther ethnic group
201187.123.17.131.70.95
202182.823.918.912.322.04

Prisoners per 100,000 residents aged 15 and over of the same ethnic group (30 June 2011, 2021, 2026)

Each line is the source that counts most for that group.

  • White
  • Black or Black British
  • Asian or Asian British
  • Mixed
  • Other ethnic group
02505007501,0002011201520202026
GroupFigureYearTrust
White149.04202663 Moderate
Black or Black British554.91202663 Moderate
Asian or Asian British168.19202663 Moderate
Mixed423.08202663 Moderate
Other ethnic group210.76202663 Moderate
Show every year as a table
YearWhiteBlack or Black BritishAsian or Asian BritishMixedOther ethnic group
2011153.11802.73214.47444.46118
2021136.87539.16150.61381.46116.77
2026149.04554.91168.19423.08210.76

Homicide victims per million people per year, by ethnic appearance (three-year averages)

Each line is the source that counts most for that group.

  • White
  • Black, Black British, Black Welsh, Caribbean or African
02040602019202020212022202320242025
GroupFigureYearTrust
White7.74202571 Moderate
Black, Black British, Black Welsh, Caribbean or African36.8202571 Moderate
Asian, Asian British or Asian Welsh10.62202571 Moderate
Mixed or Multiple ethnic groups8.04202571 Moderate
Other ethnic group9.9202571 Moderate
Show every year as a table
YearWhiteBlack, Black British, Black Welsh, Caribbean or African
201910.0437.91
20228.7140.26
20257.7436.8

Convicted principal homicide suspects per million people per year, by ethnic appearance (April 2022-March 2025)

Each line is the source that counts most for that group.

GroupFigureYearTrust
White3.97202571 Moderate
Black, Black British, Black Welsh, Caribbean or African31.68202571 Moderate
Asian, Asian British or Asian Welsh4.9202571 Moderate
Mixed or Multiple ethnic groups6.98202571 Moderate
Other ethnic group5.57202571 Moderate
Show every year as a table
Year

Records with ethnicity not stated or unknown, % of all records

Each line is the source that counts most for that group.

  • Arrests: self-defined ethnicity not stated
  • Indictable convictions: ethnicity unknown
  • Prison population: not stated or not recorded
01020304020052010201520202026
GroupFigureYearTrust
Arrests: self-defined ethnicity not stated11.13202663 Moderate
Indictable convictions: ethnicity unknown28.39202563 Moderate
Prison population: not stated or not recorded0.88202663 Moderate
Show every year as a table
YearArrests: self-defined ethnicity not statedIndictable convictions: ethnicity unknownPrison population: not stated or not recorded
2005––0.8
2006––0.86
20077.31–0.73
20084.06–0.71
20092.27–0.29
20102.399.682.23
20111.899.322.48
20121.8810.342.84
20132.0911.151.9
20142.512.561.26
20152.5814.560.71
20165.4117.540.65
20176.7221.460.7
20186.7123.870.6
201911.8325.590.63
202015.1324.710.71
202115.0226.280.79
202214.5325.821.02
202312.6531.740.99
202417.431.991.1
202515.9928.391.16
202611.13–0.88

About this data

Every source on this page (9)

Who is behind the sources

  • 65% Police or justice agency
  • 35% Statistics office
  • Other government body: shown, not counted
Each publisher, its funding and its own description
  • Home Office (UK) · Police or justice agencyFunding: UK government budget; it is a ministerial department. The Office for National Statistics publishes its police-recorded crime figures. sourceIn its own words: “The Home Office plays a fundamental role in the security and economic prosperity of the UK.” source, retrieved 2026-10-27
  • Ministry of Justice (UK) · Police or justice agencyFunding: UK government budget; it is a ministerial department. sourceIn its own words: “We work to protect and advance the principles of justice.” source, retrieved 2026-10-27
  • Office for National Statistics (ONS) · Statistics officeFunding: UK government budget; ONS is the executive office of the UK Statistics Authority, a non-ministerial department. sourceIn its own words: “Our main responsibilities are collecting, analysing and disseminating statistics about the UK's economy, society and population.” source, retrieved 2026-10-27
  • Race Disparity Unit, Cabinet Office (UK) · Other government body, shown but not counted hereFunding: UK government budget; a unit within the Cabinet Office, now called the Race Equality Unit. The arrest figures it publishes come from the Home Office. sourceIn its own words: “We collect, analyse and publish government data on the experiences of people from different ethnic backgrounds.” source, retrieved 2026-10-27
  • Ministry of Justice, Criminal Justice System statistics quarterly: December 2025, Outcomes by Offence data (CSV)Ministry of Justice (UK) (police or justice agency) · Police and justice recordsUsed for 11 numbers; counts toward 11. Years 2025. Retrieved 2026-10-27.Data (zip file)
  • Ministry of Justice, Offender management statistics: annual prison population 2026 (Table 1.A.21) and 2015 (Table A1.8)Ministry of Justice (UK) (police or justice agency) · Police and justice recordsUsed for 6 numbers; counts toward 6. Years 2026. Retrieved 2026-10-27.Data (spreadsheet)
  • Home Office, Police powers and procedures: arrests ethnic disparity time series, year ending March 2026Home Office (UK) (police or justice agency) · Police and justice recordsUsed for 5 numbers; counts toward 5. Years 2026. Retrieved 2026-10-27.Data (spreadsheet)
  • ONS Census 2021 (ethnic group by age, custom dataset) and Census 2011 (DC2101EW), England and WalesOffice for National Statistics (ONS) (statistics office) · Population registerUsed for 5 numbers; counts toward 5. Years 2021. Retrieved 2026-10-27.Open the publisher's pageData (API)
  • Ministry of Justice prison population (Table 1.A.21; 2015 Table A1.8) divided by ONS census populationMinistry of Justice (UK) (police or justice agency) · Police and justice recordsUsed for 5 numbers; counts toward 5. Years 2026. Retrieved 2026-10-27.Data (spreadsheet)
  • ONS, Homicide in England and Wales: appendix tables, year ending March 2025, Worksheet 7Office for National Statistics (ONS) (statistics office) · Police and justice recordsUsed for 5 numbers; counts toward 5. Years 2025. Retrieved 2026-10-26.Data (Excel file)
  • ONS, Homicide in England and Wales: appendix tables, year ending March 2025, Table 32Office for National Statistics (ONS) (statistics office) · Police and justice recordsUsed for 5 numbers; counts toward 5. Years 2025. Retrieved 2026-10-26.Data (Excel file)
  • Home Office, Police powers and procedures: arrest summary tables, year ending March 2026, Table A_01cHome Office (UK) (police or justice agency) · Police and justice recordsUsed for 1 number; counts toward 1. Years 2026. Retrieved 2026-10-27.Data (spreadsheet)
  • Ethnicity facts and figures (Cabinet Office), Arrests: data files 2006 to 2020Race Disparity Unit, Cabinet Office (UK) (other government body) · Police and justice recordsUsed for 5 numbers; counts toward 0 (shown only). Years 2020. Retrieved 2026-10-27.Not counted where: it republishes Home Office.Open the publisher's page
Why the numbers differ (5)
  • Self-defined and officer-observed ethnicity differ. In 2025/26, officers recorded no ethnicity for 2.76% of arrests, against 11.13% with no self-defined ethnicity. Officer-observed recording has no Mixed group: the Home Office notes people of mixed ethnicity are recorded under other groups such as Black or Asian. Officer-observed shares among known records were 76.6% White, 11.76% Black, 9.8% Asian and 1.84% Other (Table A_01d). (source, retrieved 2026-10-27)
  • The census base changed. Rates up to 2019/20 divide by 2011 census populations, rates from 2020/21 by 2021 census populations. The Black group grew from 1,864,890 in the 2011 census to 2,409,272 in the 2021 census, so the Black rate fell from 31.4 in 2019/20 to 22.62 in 2020/21 partly because the denominator changed. Ethnicity facts and figures says the two periods are not comparable. (source, retrieved 2026-10-27)
  • Stages and counting units differ. Arrests count each arrest of a person for an unconnected offence in a financial year and are divided by residents of all ages; convictions count people convicted in a calendar year by their principal offence, only for indictable offences, and are compared with residents aged 10 and over; the prison figure is a snapshot on 30 June, so people serving long sentences weigh more than people serving short ones. (source, retrieved 2026-10-27)
  • Missing ethnicity is not spread evenly. MoJ restricts court figures to indictable offences because around three quarters of ethnicity information for all offences at magistrates' courts is missing, mostly Single Justice Procedure cases. Even for indictable offences, ethnicity was unknown for 28.39% of people convicted in 2025. In November 2022 the Metropolitan Police changed data system; the share of its arrest records without self-defined ethnicity rose from 4% in 2021/22 to 69% in 2023/24 and was 34% in 2025/26 (Home Office note 14). (source, retrieved 2026-10-27)
  • Homicide figures use ethnic appearance as recorded by the police, while the census population and the arrest and court figures use self-identified ethnicity. (source, retrieved 2026-10-26)
Common mistakes (4)
  • Reading shares among people with a recorded ethnicity as shares of everyone. 28.39% of people convicted of indictable offences in 2025 and 11.13% of arrests in 2025/26 have no recorded ethnicity, and MoJ and the Home Office leave them out of the shares. (source, retrieved 2026-10-27)
  • Comparing conviction shares with the population of all ages. Children under 10 cannot be convicted, and the ethnic groups have different age profiles: the Mixed group is 2.88% of all residents but 2.32% of residents aged 10 and over (Census 2021). (source, retrieved 2026-10-27)
  • Reading a difference between groups as its cause. MoJ states that for most of its report no controls are applied for income, geography, offence mix or offender history, so it is not possible to say what share of a difference is attributable to ethnicity. (source, retrieved 2026-10-27)
  • Joining the arrest rates before and after 2020/21 into one trend: they use different census populations. (source, retrieved 2026-10-27)
Context (5)
Sources considered and not used (7)
  • MoJ Statistics on Ethnicity and the Criminal Justice System 2024, data tables: The ODS tables have no convictions or prison shares by year beyond five years; the report takes its court figures from the Criminal Justice Statistics Outcomes by Offence data, which is used here directly. (source, retrieved 2026-10-27)
  • Convictions for summary offences by ethnicity: Ethnicity is missing for around three quarters of all magistrates' court cases (MoJ), so only indictable offences are shown. (source, retrieved 2026-10-27)
  • Convictions per 100,000 people in each group as a headline number: MoJ publishes no such rate, and with more than a quarter of convictions lacking ethnicity a computed rate would understate every group by an unknown amount; the floor value is given in each conviction line's notes. (source, retrieved 2026-10-27)
  • Prison population by ethnicity before 2005: Most records used 1991 census ethnic codes until 2004; earlier years are only in older publications. (source, retrieved 2026-10-27)
  • Arrest rates for England and Wales including Derbyshire, 2020/21 to 2023/24: Published by the Home Office, but not for the two latest years; the series excluding Derbyshire is used so all years match. (source, retrieved 2026-10-27)
  • Arrests excluding the Metropolitan Police: The Home Office publishes this to reduce the effect of the Metropolitan Police recording change; not shown as a separate line to keep one Home Office series, but it is in the same time series file. (source, retrieved 2026-10-27)
  • Crime Survey for England and Wales victimisation by ethnicity; stop and search by ethnicity: They measure victims and police stops, not people arrested, convicted or imprisoned. (source, retrieved 2026-10-27)
Download, compare, cite and report an error

Sources

Loading…