Matt Ashby

@mattashby.com

2005 Followers 337 Following

I help people use data to reduce crime. Associate Professor, Crime Science, UCL. Former police officer. 🌐: mattashby.com

Map of England showing the percentage of adults in each ceremonial county who say they live in “the North of England”, based on a YouGov Profiles survey of 45,856 adults in England conducted in October 2025. Counties are shaded from light grey (0–10%) through progressively darker pinks and reds to dark maroon (81% or more). The map shows a clear north–south gradient. All counties in the far north of England are dark red or maroon, indicating that more than 70% of residents, and in several counties more than 80%, identify as living in the North. High levels extend across Northumberland, Tyne and Wear, County Durham, Cumbria, North and West Yorkshire, Greater Manchester, Merseyside and much of Lancashire. The strongest transition occurs across the Midlands. Northern parts of Derbyshire, Nottinghamshire, Staffordshire and Shropshire have at least half of residents identifying as living in the North, while neighbouring counties further south have substantially lower proportions. Counties such as Cheshire, Derbyshire, Nottinghamshire, Staffordshire and Shropshire form the main boundary zone, with values generally between 41% and 70%. South of this transition, most counties are pale pink or light grey, indicating that fewer than 30% of residents identify as living in the North, with many southern counties below 10%. An annotation highlights that “At least half of those who live in the northern parts of Derbyshire, Nottinghamshire, Shropshire and Staffordshire say they live in the North.” The source is YouGov Profiles, October 2025.
Alt text: Horizontal stacked bar chart titled “Support and opposition to police uses of AI”. Respondents were asked: “To what extent would you support or oppose the police using AI for each of the following purposes?” Each bar is divided into five categories: strongly oppose (dark burgundy), somewhat oppose (pink), neither (grey), somewhat support (light blue), and strongly support (green). Support is highest for fast translation for non-English speakers (66% support) and safeguarding vulnerable people faster (64% support, 17% neutral, 16% oppose). Around six in ten support using AI for predicting where police resources are needed (63%), linking related crimes (64%), identifying serious or repeat offenders and victims (63%), assisting with cold cases (62%), detecting internal police misconduct (62%), automating record-keeping and administration (61%), and managing deployments during busy periods (57%). Support is weaker for reducing civilian staff to fund frontline policing (45% support, 25% neutral, 26% oppose), predicting which individuals will offend (44% support, 24% neutral, 27% oppose), and assisting judges and jurors (43% support, 23% neutral, 30% oppose), the latter attracting the greatest opposition. Overall, support exceeds opposition for every proposed use.
Part of a presentation slide saying “Focus groups: Leeds, Lancaster, Lichfield, London”
Sunset over a broad tidal shoreline at low tide. In the foreground, wet sand and scattered dark pebbles reflect the warm light of the setting sun. Bands of exposed mudflats and shallow pools stretch across the middle distance, creating horizontal layers between the beach and the sea. The sun sits just above the horizon near the centre of the image, casting a narrow orange reflection across the water and wet sand. Beyond the sea, a range of low, dark hills and mountains forms a silhouette against the sky. The sky transitions from bright orange near the horizon to pale peach higher up, with only a few small clouds visible. The overall scene is calm and expansive, with strong reflections and contrasting dark shoreline textures.
Slope chart comparing the ten most commonly spoken main languages other than English in England and Wales in the 2011 and 2021 censuses. Languages are ranked by position in the top ten, with lines showing changes in rank between years. Polish remained the most common language in both years, with 612,000 speakers in 2021. Romanian showed the largest rise, moving from outside the 2011 top ten to second place in 2021 with 472,000 speakers. Panjabi and Urdu ranked third and fourth in 2021, followed by Portuguese, Spanish, Arabic, Bengali, Gujarati and Italian. Several languages that were in the 2011 top ten fell in rank by 2021, while Portuguese, Spanish, Romanian and Italian moved into the top ten. The chart emphasises changes in ranking rather than changes in the number of speakers, with connecting lines linking each language’s position in 2011 and 2021.
Map of Greater London showing the most common non-English language spoken in each borough, based on the 2021 Census. English is the most common language everywhere, so the map displays the second most common language. Distinct colours indicate different languages. Romanian dominates much of outer north-east and east London, including Barnet, Harrow, Waltham Forest, Redbridge, Barking and Dagenham, Havering and Bexley. Turkish is most common in Enfield, Haringey and Hackney. Spanish is most common in several central and south-west boroughs, including Islington, Wandsworth, Lambeth, Lewisham, Richmond, Southwark and the City of London. Panjabi is most common in Hillingdon, Ealing and Hounslow. Bengali is most common in Tower Hamlets and Newham. Other boroughs show French (Camden, Hammersmith and Fulham, Kensington and Chelsea), Gujarati (Brent), Arabic (Westminster), Nepali (Greenwich), Polish (Merton and Croydon), Portuguese (Bromley) and Tamil (Kingston and Sutton).
Map of Greater London showing the most common non-English language spoken in each borough, based on the 2021 Census. English is the most common language everywhere, so the map displays the second most common language. Distinct colours indicate different languages. Romanian dominates much of outer north-east and east London, including Barnet, Harrow, Waltham Forest, Redbridge, Barking and Dagenham, Havering and Bexley. Turkish is most common in Enfield, Haringey and Hackney. Spanish is most common in several central and south-west boroughs, including Islington, Wandsworth, Lambeth, Lewisham, Richmond, Southwark and the City of London. Panjabi is most common in Hillingdon, Ealing and Hounslow. Bengali is most common in Tower Hamlets and Newham. Other boroughs show French (Camden, Hammersmith and Fulham, Kensington and Chelsea), Gujarati (Brent), Arabic (Westminster), Nepali (Greenwich), Polish (Merton and Croydon), Portuguese (Bromley) and Tamil (Kingston and Sutton).
Map of Greater London showing the most common non-English language spoken in each borough, based on the 2021 Census. English is the most common language everywhere, so the map displays the second most common language. Distinct colours indicate different languages. Romanian dominates much of outer north-east and east London, including Barnet, Harrow, Waltham Forest, Redbridge, Barking and Dagenham, Havering and Bexley. Turkish is most common in Enfield, Haringey and Hackney. Spanish is most common in several central and south-west boroughs, including Islington, Wandsworth, Lambeth, Lewisham, Richmond, Southwark and the City of London. Panjabi is most common in Hillingdon, Ealing and Hounslow. Bengali is most common in Tower Hamlets and Newham. Other boroughs show French (Camden, Hammersmith and Fulham, Kensington and Chelsea), Gujarati (Brent), Arabic (Westminster), Nepali (Greenwich), Polish (Merton and Croydon), Portuguese (Bromley) and Tamil (Kingston and Sutton).
Screenshot of a social media post from Sky News. The post headline reads, “Vehicles set alight in Belfast as protesters take to the streets.” Below the headline is a split image showing two scenes of vehicles on fire. On the left, a large vehicle burns intensely on a city street, with bright orange flames and thick black smoke rising high above nearby buildings. On the right, another vehicle burns in a residential area, surrounded by a crowd of people gathered nearby. The linked article preview repeats the headline and mentions a man charged with attempted murder after a knife attack.
A presentation slide with the text “The central question: How can police organisations be ethical, equitable and efficient at the same time?”
The Elements of CRIMEFACE Crime reduction/detection outcomes Reactionary outcomes other than crime Implementation considerations Misuses/potential backfire effects Economics Future proof-ability Applicability Contextual differences Ethical considerations and sustainability
A podcast listing showing an episode of the podcast Analyst Talk featuring Scott Keay is 1 hour 41 minutes long
A crowd at a football match, with a banner saying UEFA Women’s Champions League
A conference panel including Professor Emeritus Gloria Laycock, UCL Jill Dando Institute of Security and Crime Science, Matt Twist, Metropolitan Police Service Assistant Commissioner, Sir Denis O'Connor, former Chief Inspector of Constabulary, chaired by Dame Sara Thornton, Chair of Trustees, The Police Foundation
Former Home Secretary David Blunkett, former New York City Police Commissioner Bill Bratton and Director of the Police Foundation Blair Gibbs in a panel discussion on a stage
Line chart showing male larceny-theft arrest rates per 100,000 by age group for 2005, 2015, and 2025. Across all three years, arrest rates are highest among teenagers and young adults, then decline steadily with age, but the shape and level of the curve change over time. In 2005, rates peak sharply in the late teens (around ages 17–19) at very high levels (roughly 1,400–1,500), then drop quickly through the 20s and continue declining into older ages. In 2015, the pattern is similar but less extreme: the teenage peak is lower (around 1,100–1,200), and rates in the 20s are relatively higher than in 2005, producing a more gradual decline. By 2025, the distribution is flatter and shifted toward older ages. The teenage peak is much lower (around 400), while rates in the late 20s to mid-30s are comparable to or slightly higher than those for teenagers, creating a broad plateau rather than a sharp peak. From about age 40 onward, all three years show similar declining patterns, with relatively small differences by older age. Overall, arrest rates have fallen substantially among younger males and become more evenly distributed across early and middle adulthood.
Figure 3 bar chart showing proportion of time spent on specific crime incident activities by type. “Violence with injury” is highest at 14.7%, followed by “Violence without injury” (8.3%) and “Other sexual offences” (8.1%). Mid-range categories include rape (7.2%), homicide (6.4%), trafficking of drugs (5.6%), and vehicle offences (4.8%). Several categories cluster around 3–4%, including stalking and harassment, public order offences, and fraud/forgery. Lower proportions include possession of weapons (2.7%), shoplifting (2.4%), robbery (2.2%), and burglary business/community (1.5%). The smallest shares are theft from the person (0.9%), arson (0.7%), and bicycle theft (0.2%). Overall, time is concentrated in violent offences.
A tranquil rural scene shows a narrow stream flowing through a neatly maintained garden beside a cottage. On the left, a blue-painted house with a wooden door and chimney sits behind a small gravel area, where a bench and a British flag are visible. The garden features bright green grass edged with spring flowers, including clusters of yellow daffodils. A tall, conical shrub stands near the house, while a weeping willow drapes its branches over the water. The stream runs gently over stones toward a small stone bridge in the background, with trees and hedges lining both banks under a cloudy sky.
Three side-by-side line charts show long-term trends in non-sexual violent crime in England and Wales from 1993 to 2023, measured as crimes per 1,000 people aged 16+. Categories are serious injury, minor injury, and no injury. All three series decline substantially over time. Serious injury falls steadily from the late 1990s to a low level by 2023. Minor injury shows a sharp drop in the early 2000s, a brief plateau around 2010, then further decline. No injury starts highest, decreases gradually, with a small mid-period flattening, then continues downward. Data are three-year moving averages from the Crime Survey for England and Wales.
We are very pleased to announce we will: Increase paid paternity/partner's leave from 4 to 12 weeks from 1 May this year. Increase paid maternity and adoption leave from 18 weeks to 22 weeks from 1 May this year to ensure all parents have increased time off at this important time with less financial worry. Introduce a paid leave period of two weeks for parents who experience pregnancy loss before 24 weeks of pregnancy (a week of paid leave to partners will also be provided). Partner with an external provider to introduce backup care for all staff where their regular caring provision breaks down last minute, disrupting their normal working pattern.
Sunbeams stream through thick clouds in a bright sky, illuminating a rural landscape with the silhouette of a lone leafless tree on the horizon.
The cover of a book titled (in German) ‘Forensics in 30 seconds: crime investigation through science’
We, the Editor and Publishers of Journal of Criminal Justice Education, have retracted the following book review: Kresnanda Annas, G. (2025). Organising White-Collar and Corporate Crimes: By Nicholas Lord and Michael Levi, New York: Routledge, Published 2025, 218 p., 19 B/W Illustrations, £35.99 (eBook), ISBN: 9781315100227. Journal of Criminal Justice Education, 30 April 2025, DOI: 10.1080/10511253.2025.2500310. Following publication, a third party raised concerns regarding the integrity of the review. Upon investigation and following correspondence with the author, the Journal and Publisher confirm that this book review contains such extensive inaccuracies that it neither describes nor evaluates the book it claims to review. We have been informed in our decision-making by our own publicly available editorial policies and COPE guidelines. The retracted book review will remain online to maintain the scholarly record, but it will be digitally watermarked on each page as ‘Retracted’.
A large neoclassical stone building at night with the words UCL200 projected onto it
Betteridge's law of headlines is an adage that states: "Any headline that ends in a question mark can be answered by the word no." It is based on the assumption that if the publishers were confident that the answer was yes, they would have presented it as an assertion; by presenting it as a question, they are not accountable for whether it is correct or not.
A fast-flowing river beneath an overhanging tree that is covered in bright green moss
Snowdrops at the side of a road
Bar chart showing the percentage of homicides believed to be gang-related or organized crime group (OCG)-related in England and Wales, categorized by circumstance. The x-axis shows percentages from 0% to 40%, and the y-axis lists homicide circumstances. Dark blue bars represent gang-related cases, and light blue bars represent OCG-related cases. “Long running disputes” accounts for the highest percentage in both categories, with nearly 50% gang-related and about 27% OCG-related. “In the course of the suspect committing another crime” is the second highest for OCG-related homicides, while “non-domestic fights, brawls, etc.” and “other known circumstances” are the next most common for gang-related cases. Other categories like “domestic abuse,” “sexual motive,” and “terrorist attack” show negligible or no association with gang or OCG involvement.
The cover of a report titled 'Patterns and trends in homicide in England and Wales and their implications for policing: a problem profile'
Bar chart comparing domestic homicide victimisation rates per million people in England and Wales by ethnicity and sex. The chart is split horizontally: light blue bars show female victims, dark blue bars show male victims. Domestic-homicide risk is highest for Black women (7 homicides per million people), Black men (4.5 homicides per million) and Asian women and women of other ethnicities (4 per million). Looking at the number of victims in each category, the highest average annual number of victims is among White women (60 victims), followed by Asian (8), Black (7), Other (2), and Mixed/Multiple (1). Among males, White victims also have the highest rate at 29, followed by Asian (5), Black (4), Other (1), and Mixed/Multiple (0). The x-axis shows victimisation rate per million people; the y-axis lists ethnicities, grouped by sex.
The cover of a report titled 'Patterns and trends in homicide in England and Wales and their implications for policing: a problem profile'
Bar chart comparing domestic homicide victimisation rates per million people in England and Wales by ethnicity and sex. The chart is split horizontally: light blue bars show female victims, dark blue bars show male victims. Domestic-homicide risk is highest for Black women (7 homicides per million people), Black men (4.5 homicides per million) and Asian women and women of other ethnicities (4 per million). Looking at the number of victims in each category, the highest average annual number of victims is among White women (60 victims), followed by Asian (8), Black (7), Other (2), and Mixed/Multiple (1). Among males, White victims also have the highest rate at 29, followed by Asian (5), Black (4), Other (1), and Mixed/Multiple (0). The x-axis shows victimisation rate per million people; the y-axis lists ethnicities, grouped by sex. A note explains that 97.5% of homicides had known victim sex and ethnicity, and population data refers to those aged 16+ per the Home Office definition of domestic homicide. Data source is the Home Office Homicide Index.
Bar chart showing the percentage distribution of homicide victims by age, sex, and location of the incident in England and Wales. The chart is divided into three columns by location—“other location,” “street, open space, etc.,” and “residence”—and two rows by sex—female victims (top) and male victims (bottom). For both sexes, homicides in residences are most common among older age groups. Among females, most victims in all age groups were killed in residences. Among males, younger victims (especially under 40) are more likely to be killed in streets or open spaces. Male victims show a higher percentage in “other locations” across most age groups than females. Data source: Home Office Homicide Index.
Horizontal bar chart showing the percentage of homicides by location and the suspect’s relationship to the victim, based on data from the Home Office Homicide Index. The categories of relationship are color-coded: dark blue for partner/ex-partner, light blue for family member, pink for acquaintance, and dark red for stranger. The most common location for homicides is a house or dwelling, with a broad distribution across all relationship types, notably including a large proportion by partners/ex-partners and family members. The second most common location is street or open space, dominated by stranger-related homicides. Other locations include “other location”, “licensed premises”, “hotel/hostel”, and “unknown”, each contributing a much smaller share. Stranger-related homicides are prevalent in public spaces, while domestic settings see more cases involving partners or family. The data represents the 67.6% of homicides where the suspect–victim relationship was known.
A bar chart shows the relationship between neighbourhood deprivation and homicide rates in England. The x-axis represents the population of England arranged by level of deprivation, from 0% (most deprived) to 100% (least deprived). Each bar represents 1% of the population and shows the annual number of homicides per million residents in that group. The y-axis ranges from 0 to 35 homicides per million. Homicide rates are highest in the most-deprived areas on the left side of the chart, with most-deprived percentile showing a rate over 30 homicides per million. The rates generally decline as deprivation decreases, although there are some fluctuations. The right side of the chart, representing less-deprived areas, shows consistently lower homicide rates, mostly below 10 homicides per million. Above the chart, a row of percentages indicates the share of total homicides that occurred in each decile of deprivation. These are: 23%, 16%, 14%, 12%, 10%, 8%, 7%, 5%, 4%, and 2%.
A hexagonal heat map of London shows the density of homicides based on data from the Home Office Homicide Index. The map uses a gradient from light beige to dark red-orange to indicate areas of low to high homicide density. Central and southeastern parts of London, including areas around Stockwell, Warwick Avenue, and Newbury Park, show the highest concentrations, marked by darker red-orange hexagons. Other named areas include Turnpike Lane and Selhurst, which show moderate to lower densities. The map features borough boundaries and a scale bar indicating 8 km for reference. A legend at the bottom left explains the color gradient, and a note beneath it states that the data covers 94.4% of cases with known offence locations. The map emphasizes geographical variation in homicide frequency across Greater London.
Scatter plot showing the annual homicide victimisation rate per 100,000 residents across local authorities in England and Wales, grouped by region. Each dot represents a local authority, with red circles for mainly urban areas and blue for mainly rural. The x-axis shows homicide rates ranging from 0 to over 5; the y-axis lists regions like London, West Midlands, and East of England. Black bars represent the regional average homicide rate, while a dashed vertical line shows the national average. London shows many urban authorities with rates above 1. Some outliers with high rates include Lambeth, Greenwich, and Boston. South East and East of England have clusters of low-rate rural areas. The chart visualizes both the variability within regions and the urban-rural divide in homicide rates. Sourced from the Home Office Homicide Index, based on 94.4% of cases with known offence locations. Legend clarifies circle colors and rate markers.
Side-by-side choropleth maps of England and Wales show homicide rates by police jurisdiction, separated by victim sex. The left map displays male homicide victims with a national rate of 45.4 per million residents. Darker shades indicate higher rates, with Cleveland (northeast) having the highest, followed by West Midlands Police (WMP), the Metropolitan Police Service (MPS), Greater Manchester Police (GMP), and Lincolnshire. The right map shows female homicide victims with a national rate of 16.6 homicides per million residents. Dyfed-Powys in Wales has the highest rate, followed by Warwickshire, Norfolk, Staffordshire, and West Mercia. Both maps use varying shades of blue to represent homicide rates per million, with a darker hue indicating a higher rate. A note clarifies that victims from the Grays lorry incident were excluded. The maps are sourced from the Home Office Homicide Index and based on data where victim sex was known, covering 99.9% of all homicides recorded.
A choropleth map of England and Wales showing homicide rates per million residents by police force area, based on Home Office Homicide Index data. The map uses a gradient from light (lower rates) to dark blue (higher rates). Each police force region is outlined and numbered if it’s among the top ten for homicide rates. Cleveland ranks highest at 49.2 homicides per million, followed by Greater Manchester (46.7), West Midlands (45.6), and the Metropolitan Police area (45.0). Other top areas include Dyfed-Powys, Lincolnshire, West Mercia, Lancashire, and Warwickshire. A legend on the right indicates the color scale ranging from 10 to over 50 homicides per million. The map excludes 39 victims from the Grays lorry incident. Lighter-colored areas represent lower homicide rates, mainly in southern and rural regions, while darker areas are concentrated in some urban and northern regions.
The cover of a report titled 'Patterns and trends in homicide in England and Wales and their implications for policing: a problem profile'
Side-by-side choropleth maps of England and Wales show homicide rates by police jurisdiction, separated by victim sex. The left map displays male homicide victims with a national rate of 45.4 per million residents. Darker shades indicate higher rates, with Cleveland (northeast) having the highest, followed by West Midlands Police (WMP), the Metropolitan Police Service (MPS), Greater Manchester Police (GMP), and Lincolnshire. The right map shows female homicide victims with a national rate of 16.6 homicides per million residents. Dyfed-Powys in Wales has the highest rate, followed by Warwickshire, Norfolk, Staffordshire, and West Mercia. Both maps use varying shades of blue to represent homicide rates per million, with a darker hue indicating a higher rate. A note clarifies that victims from the Grays lorry incident were excluded. The maps are sourced from the Home Office Homicide Index and based on data where victim sex was known, covering 99.9% of all homicides recorded.
Bar chart showing proportions of homicide methods across eight sub-types of circumstances in England and Wales, based on the Home Office Homicide Index. Each sub-type is displayed in its own chart with coloured bars representing the proportion of methods used within that sub-type and grey bars showing overall proportions across all homicides. The sub-types are: (1) Female victims of domestic homicide aged 16+, (2) Male victims of domestic homicide aged 16+, (3) Male victims under 25 killed in public spaces, (4) Female victims, non-domestic, aged 16+, (5) Male victims, non-domestic, killed in public spaces, (6) Male victims, non-domestic, killed in homes, (7) Victims under 16 killed in non-public places, and (8) Other victims. Methods include sharp instrument, hitting/kicking, strangulation, blunt instrument, shooting, other, and not known. Further information in linked blog post.
A grouped bar chart shows the percentage of homicide victims by age and sex for different methods of killing in England and Wales, based on data from the Home Office Homicide Index. Age is grouped in ten-year categories along the x-axis, while the y-axis shows percentage of victims within each age-sex group. Separate horizontal panels compare female (top) and male (bottom) victims. Each panel includes six sub-charts for homicide methods: sharp instrument, hitting/kicking, blunt instrument, strangulation, shooting, and other. Male victims are more often killed by sharp instruments, especially younger males. Older males are more often killed by hitting/kicking. Female victims are more evenly distributed across methods, with strangulation showing a higher share in younger women. “Other” methods are more frequent in the very young and elderly for both sexes. Chart notes indicate that 93.4% of cases had known victim sex and offence method.
Table comparing firearm and knife homicide rates per million people in four developed, English-speaking countries. The table lists England and Wales, Australia, Canada, and the United States, along with the year of data and homicide rates for firearms and knives. For 2022–23, England and Wales had 0.5 firearm and 4.1 knife homicides per million people; Australia had a rate of 1.5 firearm and 3.6 knife homicides. In 2023, Canada recorded a rate of 9.9 firearm and 8.0 knife homicides, while the U.S. had a rate of 59.1 firearm and 6.6 knife homicides. Sources include national statistical agencies from each country, with linked references listed below the table.
Bar chart comparing firearms homicide rates in various countries to England and Wales, where the rate is set as the baseline (1.0). The United States has the highest rate, at 85.9 times higher. Chile (64.5 times), Canada (14.3 times), and Sweden (9.8 times) follow. Most other countries, including Germany, Denmark, and Australia, range from 1.9 to 0.8. Japan, Hong Kong, and South Korea had zero or near-zero rates. Data includes countries with over 5 million population and available firearms homicide statistics since 2016. Bars are shaded blue and listed left to right in decreasing order of homicide rate per million residents.
Bar chart showing average annual number of homicide victims in England and Wales by method and gender, based on Home Office Homicide Index data. The most common method is sharp instruments, with around 260 victims, predominantly male. Next is hitting, kicking, etc., with around 109 victims. Strangulation, blunt instruments, and other methods follow with 30–50 victims each. Shooting, use of vehicles, poison, and burning have fewer victims. Drowning and explosions are least common. Gender is color-coded: pink for male and dark teal for female victims. A category labeled “not known” includes 41 victims where the method was unclear.
Cover of a report called 'Patterns and trends in homicide in England and Wales and their implications for policing: a problem profile'
A heatmap-style chart shows percentages of homicide victims by their ethnicity and the ethnicity of the suspect, based on data from the UK Home Office Homicide Index. The chart has victim ethnicities listed on the vertical axis (White, Black, Asian, Mixed/Multiple, Other) and suspect ethnicities on the horizontal axis with the same categories. Each cell contains a percentage and is shaded according to the proportion of total homicide victims, with darker green indicating higher values. The most prominent cell shows that 51% of homicide victims were White individuals killed by White suspects. Other significant cells include Black victims with Black suspects (10%) and Other victims with White suspects (12%). Many cells have values near or at 0%. A vertical color scale on the right indicates shading from 0% (light) to 60% (dark). The data reflects only the 56.8% of cases where both suspect and victim ethnicities were known.
This grouped bar chart shows the percentage of homicide suspects’ relationships to victims in England and Wales, broken down by victim age (in ten-year categories) and sex (female on top, male on bottom). Four relationship types are represented across the chart: partner/ex-partner, family member, acquaintance, and stranger. Among female victims, suspects are most often current or former partners, especially for victims aged 20–49. In contrast, for male victims, suspects are more commonly acquaintances or strangers across all age groups. Family members are most likely to be suspects in homicides of the very young (ages 0–9) and elderly (70+), especially for male victims. For female victims, acquaintance and stranger suspects are less common overall but rise slightly in older age groups. Among male victims, partner/ex-partner suspects account for a very small proportion across all ages. The chart highlights the gendered nature of homicide suspect relationships, particularly the high …
Bar chart showing annual homicide suspect rates per million people by age and ethnicity in England and Wales. The chart is divided into five panels: All suspects, Asian, Black, Mixed/Multiple, and White. Bars are color-coded: dark red indicates a lower or equal homicide rate compared to the total population of the same age, while bright red indicates a higher rate. A stepped line in each panel shows the overall homicide offending rate across all ethnicities for reference. Black suspects, especially aged 20–30, show significantly higher homicide suspect rates compared to other groups, with the highest peak exceeding 80 per million. Other groups peak around ages 20–30 but at lower rates. Asian, White, and All suspects show more moderate rates, while Mixed/Multiple suspects have a small peak. The x-axis shows age in 10-year increments; the y-axis shows homicide suspect rate. Data is from the Home Office Homicide Index, covering 95.1% of known-age, known-ethnicity suspects.
Chart showing age distribution of convicted homicide offenders by victim type in England and Wales, with separate density curves for seven sub-types and an “other victims” group. Each curve shows the spread of offender ages, with a vertical line indicating the median. Female victims of domestic homicide (16+): Wide age range with a peak in offenders aged 30–40, tapering off gradually on both sides. Median around age 42. Male victims of domestic homicide (16+): Bimodal distribution with prominent peaks around ages 25 and 45. Median around age 33. Male victims under 25 killed in public spaces: Sharp peak around age 19, indicating most offenders were young. Steep drop-off after age 25. Female victims, non-domestic (16+): Flatter distribution with no strong peak, suggesting a wide range of offender ages. Median around 32. Male victims, non-domestic, killed in public spaces: Skewed toward younger offenders with a peak around age 20. Median 25. Male victims, non-domestic, killed in homes: …
Bar chart showing the number of homicide suspects in England and Wales by age, split into child (light green) and adult (dark green) categories. The chart spans ages 10 to 85 and peaks at age 19, where over 70 suspects are recorded. A sharp decline follows into the late twenties, then gradually decreases with age. Very few suspects are aged over 60. Data source is the Home Office Homicide Index, including only cases where the suspect’s age was known.
Cover of a report called 'Patterns and trends in homicide in England and Wales and their implications for policing: a problem profile'
Bar chart showing annual homicide suspect rates per million people by age and ethnicity in England and Wales. The chart is divided into five panels: All suspects, Asian, Black, Mixed/Multiple, and White. Bars are color-coded: dark red indicates a lower or equal homicide rate compared to the total population of the same age, while bright red indicates a higher rate. A stepped line in each panel shows the overall homicide offending rate across all ethnicities for reference. Black suspects, especially aged 20–30, show significantly higher homicide suspect rates compared to other groups, with the highest peak exceeding 80 per million. Other groups peak around ages 20–30 but at lower rates. Asian, White, and All suspects show more moderate rates, while Mixed/Multiple suspects have a small peak. The x-axis shows age in 10-year increments; the y-axis shows homicide suspect rate. Data is from the Home Office Homicide Index, covering 95.1% of known-age, known-ethnicity suspects.
A horizontal bar chart shows the percentage of homicides in England and Wales by circumstance and the suspect’s relationship to the victim. Categories include “Fights, brawls, etc.”, “Domestic abuse”, and “In the course of another crime,” among others. The four suspect relationship types are color-coded: partner/ex-partner (dark blue), family member (light blue), acquaintance (pink), and stranger (burgundy). “Fights, brawls, etc.” is the most common category overall, with a majority committed by acquaintances and strangers. “Domestic abuse” is the second most common, predominantly involving partners/ex-partners and family members. Homicides during other crimes and long-running disputes also rank high, largely involving strangers and acquaintances. Less frequent categories include child abuse, irrational acts, reckless acts, sexual motives, and terrorism. A significant portion of cases fall under “Unknown circumstances.” The data source is the Home Office Homicide Index, based on 67.6%
A horizontal bar chart shows the percentage of homicides in England and Wales by circumstance and the suspect’s relationship to the victim. Categories include “Fights, brawls, etc.”, “Domestic abuse”, and “In the course of another crime,” among others. The four suspect relationship types are color-coded: partner/ex-partner (dark blue), family member (light blue), acquaintance (pink), and stranger (burgundy). “Fights, brawls, etc.” is the most common category overall, with a majority committed by acquaintances and strangers. “Domestic abuse” is the second most common, predominantly involving partners/ex-partners and family members. Homicides during other crimes and long-running disputes also rank high, largely involving strangers and acquaintances. Less frequent categories include child abuse, irrational acts, reckless acts, sexual motives, and terrorism. A significant portion of cases fall under “Unknown circumstances.” The data source is the Home Office Homicide Index, based on 67.6%
Horizontal bar chart showing annual homicide victimisation rates per 100,000 people in England and Wales, broken down by age, sex, and ethnicity. The highest rates are for Black males aged 16–24 (21.5) and 25–34 (14.4), followed by Asian males 16–24 (3.2). Most victims with higher rates are young to middle-aged males, especially Black males. Females across all groups show significantly lower rates, all below 1.2. Bars are colored red for males and blue for females. Data is sourced from the Home Office Homicide Index, covering 90.6% of cases where victim demographics were known, excluding mixed/other ethnicities.
A set of eight horizontal bar charts compares annual homicide victimisation rates per million people across different ethnic groups in England and Wales, segmented by victim Home Office homicide sub-type. Each chart includes five ethnic groups: Black, Other, Asian, White, and Mixed/Multiple. Dashed lines show the average for each sub-type. Black individuals consistently show higher victimisation rates, especially among male victims under 25 killed in public (peaking near 60 per million). Data is from the Home Office Homicide Index, excluding the Grays lorry deaths.
A heatmap chart showing homicide disparity ratios in England and Wales by victim ethnicity, sex, and age group. Rows represent ethnic groups (Asian, Black, and Other) and columns represent age groups for female and male victims. The number in each cell shows the rate of homicides in of people of a particular ethnicity relative to the rate of homicides for White people of the same age and sex. The darkest red cells, indicating highest relative risk, appear for Black males aged 10–17 and 18–29, with risk ratios of ×13.4 and ×13.3 respectively. Other males aged 18–29 and Black females 65+ also show elevated risks. Cells with fewer than 5 homicides are marked with a dash (–). Source: Home Office Homicide Index. Base: 96.2% of homicides with known age, sex, and ethnicity.
Bar chart showing homicide victimisation rates per million people in England and Wales by age group and ethnicity, based on data from the Home Office Homicide Index. The chart is divided into five panels by ethnicity: All victims, Asian, Black, Mixed/Multiple, and White victims. Each panel displays vertical bars for age groups in ten-year intervals from 0 to 80+, with darker bars indicating lower or equal homicide rates compared to the total population of the same age, and red bars indicating higher rates. A stepped line overlays each panel, showing the homicide rate for all ethnicities for reference. The chart highlights significantly higher homicide rates among young Black victims at all ages, peaking at over 100 victims per million at age aged 20–29. Other ethnic groups generally show lower or comparable homicide rates to the overall population, with slight increases among certain age groups. The data is based on the 95.1% of homicides for which both age and ethnicity were known.