How serious?
499
police-recorded homicides
2025–26
Rare events can cause devastating harm. Domestic abuse and sexual violence also need separate evidence.
Explore the evidence →Headline evidence · Crime & safety · England and Wales · Year ending June 2026
Recorded knife offences fell 6%. Fraud rose 11%. Do we feel safe?
What the latest figures mean for your safety, your neighbourhood and what happens when you report a crime.
Proposed outcome: Reduce the harm crime causes and help protect people from being targeted again. What could change →
- Geography
- England and Wales
- Evidence period
- Year ending June 2026
- Evidence status
- Site calculation from published data
- Editorial review
- Reviewed 23 October 2026
The short version
England and Wales · Year ending June 2026
Changes compared with the previous year
Police-recorded
−6%
Survey estimate
+11%
Each figure is compared with the same 12 months a year earlier. Police records and survey estimates are different measures and are not added together.
Serious violence · England and Wales
Police-recorded knife offences fell 8.1% (YE Mar 2026)
48,774
recorded offences · 2025–26
8% lower
than 2024–25
28%
of recorded knife offences were in the Metropolitan Police area
Police-recorded knife or sharp instrument offences
Thousands of offences, 2010–11 to 2025–26. Two runs: recording improved before the NDQIS method began in April 2019; the Metropolitan Police corrected its recording from April 2024.
An offence count is not a count of injuries. Recording changes limit long comparisons; the chart shows the available record, with separate runs across the method change.
Budget 2026 · 28 October 2026
What the Budget means for crime & safety
3
Budget measures on this topic
£360m
Net cost, 2026–27
£10m
Net cost, 2029–30
Biggest single-year effect of each measure · red costs money, teal saves or raises it
- Further Priorities2026–27−£355m
- Tackling Misogyny2026–27−£5m
- Waste Crime2027–28−£5m
−£355m
cost in 2026–27
Further Priorities
−£5m
cost in 2026–27
Tackling Misogyny
−£5m
cost in 2027–28
Waste Crime
The Budget in 2 minutes →Is it being delivered? →
Treasury costings certified by the OBR, not measured results. Topic links are UK Facts’ classification.
The evidence
Is knife crime falling? How widespread is antisocial behaviour? What happens when a crime is reported? Start with the problems people encounter, then follow the evidence on what could change them.
England and Wales. Survey estimates describe experiences; police records describe what reaches the recorded system.
What people believe and what is measured
About eight in ten adults say crime in the country has gone up. The Crime Survey, which asks people about what happened to them, shows crime excluding fraud has fallen by a third in ten years. The two have not moved together: in 2009 84% said crime was rising when measured crime was at 10.4 million incidents; in 2015, 57% said so when it was 7.2 million. People are less negative about their own area (52% say crime there has gone up) and most feel safe walking alone after dark (78%, up from 75% ten years ago). Concern is real and survey answers are valid; they measure what people see and hear, not what happens.
What people say
What the survey measures
Years ending March; no survey estimates for 2021 and 2022. England and Wales, adults 16+. Over these years the two lines have no clear link (correlation 0.02 across 16 survey years; descriptive only).
Feel safe walking alone after dark
Percent who feel very or fairly safe. A fall in the share who feel safe would show up here; it has not.
- All adults, 201675%
- All adults, 202678%
- Women, 201662%
- Women, 202668%
- Men, 202688%
High worry about crime, 2026
People worry about fraud about twice as much as about violence.
- Fraud17%
- Violent crime9%11.5% in 2016
- Burglary8%
- Car crime7%
Police-recorded violence rose from 601,000 in 2012–13 to 1.96 million in 2025–26. Most of that is better recording, which the statistics regulator and inspectors have said, so it is not shown as a trend. Homicide, the offence least affected by recording, was 499 in 2025–26, against 711 in 2019–20, a year that includes 39 people found in a lorry in Grays, Essex.
04 · From experience to response
What happens after something goes wrong?
An incident has to reach the police before it can appear in their records. Investigation and court capacity then shape the response. These measures describe different stages; they are not a single case-tracking funnel.
8.5% of recorded crimes led to a charge or summons in 2025–26, against 15.5% in 2014–15 (a change of −7.0 percentage points). Open Crown Court cases were 80,829 at Jun 2026, 67% more than at Mar 2016.
It has to be reported
36%
of survey crime excluding fraud was reported to police (2025–26). Much of what people experience never enters police records.
It needs a response
145,886
full-time-equivalent police officers (Mar 2026). Numbers alone do not show where officers work or which skills are available.
Cases need to progress
80,829
open Crown Court cases (Jun 2026). A caseload, not a count of victims or a measure of each person’s wait.
Fraud: more cases, but not more losses
Fraud is higher than when the survey first measured it: 4.5m incidents in the year to March 2026, against 3.4m in 2016–17, and 46.6% of Crime Survey headline incidents. The change on the previous year was not statistically significant in the year to March 2026; in the year ending June 2026 the 11% rise was statistically significant. It is not the main story, because over half of it ends with the money paid back and another quarter with no loss. UK Finance reports unauthorised card fraud losses of £703m in 2025, down from £784m in 2020, while the number of cases rose. Scams where people are tricked into paying are rising: £576m, up from £421m. Fraud cost an estimated £14.4bn in 2023-24, probably an underestimate. Fraud cases are up. Losses are not rising in proportion: unauthorised card losses are lower than in 2020 while cases are higher. The survey has no fraud estimate before 2017, so a 30-year link to technology cannot be tested with official data.
54%27%19%
- Loss, fully paid back: 54.0%
- No loss (including attempts): 27.1%
- Loss, not or only part paid back: 18.9%
Where the money goes
Public order and safety cost £55.7bn in 2025–26, 1.8% of GDP. Spending fell after 2010 and has recovered. There were 145,886 police officers in March 2026, about the same as 2010 (143,734), but fewer per head of population. We have not costed options to change this.
UK public order and safety spending in real terms (2025-26 prices). It fell to £40.9bn in 2016-17 and has recovered. Police services were £31.6bn of £55.7bn in 2025-26.
- 2008-09£52.2bn
- 2010-11£49.7bn
- 2016-17£40.9bn
- 2020-21£47.3bn
- 2024-25£53.5bn
- 2025-26£55.7bn
Knife crime and antisocial behaviour: what is happening?
A knife attack can change a life in seconds. Repeated intimidation or disorder can change how people use their own neighbourhood. These are different kinds of harm, and each needs its own evidence.
What does daily life feel like?
40.9%
experienced OR witnessed local ASB
2025–26
Not a direct-victim rate. A questionnaire context change inflated part of the annual increase.
Explore the evidence →Beyond counts: the official Crime Severity Score
ONS weights police-recorded offences using sentencing information. This published measure adds a view of seriousness; it is not a complete measure of trauma, fear or financial loss, and recording changes still matter. Its release period differs from the latest crime survey.
+2.6%YE Mar 2003–YE Mar 2025England and Wales · score15.9 → 16.3YE Mar 2003–YE Mar 2025
The score moved from 15.9 (YE Mar 2003) to a low of 9.3 (YE Mar 2013), then to 16.3 (YE Mar 2025). The endpoint badge does not describe a steady rise. ONS published sentencing weights, England and Wales. Recording changes limit comparisons; do not rank neighbourhood safety with this score.Sources and data
ONS published sentencing-based severity weights; recording and offence mix affect the score. Survey fraud, unreported crime and personal harm are not fully captured. Police recording improvements limit long-run interpretation. Use the complete latest ONS vintage, never mix scoring weights across vintages.
Source dataset →| Period | score | Status |
|---|---|---|
| YE Mar 2003 | 15.889 | Outturn |
| YE Mar 2004 | 15.689 | Outturn |
| YE Mar 2005 | 14.23 | Outturn |
| YE Mar 2006 | 13.975 | Outturn |
| YE Mar 2007 | 13.314 | Outturn |
| YE Mar 2008 | 12.054 | Outturn |
| YE Mar 2009 | 11.293 | Outturn |
| YE Mar 2010 | 10.791 | Outturn |
| YE Mar 2011 | 10.354 | Outturn |
| YE Mar 2012 | 9.816 | Outturn |
| YE Mar 2013 | 9.315 | Outturn |
| YE Mar 2014 | 9.355 | Outturn |
| YE Mar 2015 | 9.962 | Outturn |
| YE Mar 2016 | 10.99 | Outturn |
| YE Mar 2017 | 12.177 | Outturn |
| YE Mar 2018 | 14.857 | Outturn |
| YE Mar 2019 | 15.8 | Outturn |
| YE Mar 2020 | 15.586 | Outturn |
| YE Mar 2021 | 13.658 | Outturn |
| YE Mar 2022 | 15.88 | Outturn |
| YE Mar 2023 | 16.355 | Outturn |
| YE Mar 2024 | 16.157 | Outturn |
| YE Mar 2025 | 16.294 | Outturn |
The threats have changed. So must the response.
Theft and violent incidents are far below their earlier peaks. But that does not tell us how many people lose money to scams today. The two timelines show why yesterday’s crime picture cannot explain every risk people face now.
Survey-based evidence
Crime Survey for England and WalesFieldwork: year to March 2026A household survey of adults aged 16 and over in England and Wales; estimates have sampling uncertainty.
Which way is crime moving?
The risks have shifted: fewer thefts and violent incidents, more fraud.
Theft, violence and criminal damage have fallen over the long term
Millions of incidents, 1981 to 2025–26. Comparable throughout; gap where the survey stopped during Covid.
1996–2026 · recorded figures.
Fraud incidents: the longer-term rise behind today’s scams
Millions of incidents, 2016–17 to 2025–26 — no comparable estimates exist before 2016–17.
Fraud incidents: no statistically significant change (YE Mar 2026)
1996–2026 · recorded figures. Shading marks years without figures: this series starts in 2017.
Theft and violence covered by the survey have fallen over the long run. Fraud is higher than when it was first measured. Police-recorded knife crime has fallen recently; recording changes limit older comparisons.
Show the workings — The risks have shifted: fewer thefts and violent incidents, more fraud.
Each chart uses its own measure and scale; survey estimates and police counts cannot be added.
Survey crime outside fraud is about four-fifths below its 1995 peak. Fraud has no comparable history before 2016–17 and is higher than when first measured. Police-recorded knife offences fell 8.1% (YE Mar 2026). Recording improvements mean the full historical series is not comparable throughout.
- Geography:
- England and Wales — A and B: adults 16+ and households; C: all recorded offences, all ages
- Measure:
- A and B: millions of incidents; C: thousands of offences
- Period:
- A: 1981 to 2025–26; B: 2016–17 to 2025–26; C: 2010–11 to 2025–26
- Status:
- Survey-based
- Calculation:
- As published. A and B from Table A1a; C from Worksheet B6 (total selected offences including homicide, including Greater Manchester). Survey years to 1999 are calendar years; from 2001–02 they are interview years ending March.
- Compare with:
- Each panel against its own earlier years only — the panels are not on a common scale and must not be compared with each other.
- Not included / breaks:
- No survey estimates for 2020–21 and 2021–22 (face-to-face interviewing stopped during Covid): the lines break there. Fraud was not measured before 2016–17. Knife offences: most forces improved recording before the NDQIS method began in April 2019, so the two runs are drawn separately; Greater Manchester undercounted before December 2017; the Metropolitan Police corrected its recording from April 2024 without revising earlier years.
Source: ONS, Crime in England and Wales: Appendix tables, year ending March 2026, Table A1a (released 23 July 2026).
View the data as a table
| Year | A: survey crime excl. fraud (m) | B: survey fraud (m) | C: knife offences (recorded) |
|---|---|---|---|
| 1981 | 11.30 | – | – |
| 1983 | 12.13 | – | – |
| 1987 | 13.65 | – | – |
| 1991 | 15.56 | – | – |
| 1993 | 18.84 | – | – |
| 1995 | 19.79 | – | – |
| 1997 | 17.17 | – | – |
| 1999 | 15.25 | – | – |
| 2001–02 | 12.77 | – | – |
| 2002–03 | 12.39 | – | – |
| 2003–04 | 11.61 | – | – |
| 2004–05 | 10.78 | – | – |
| 2005–06 | 10.86 | – | – |
| 2006–07 | 11.18 | – | – |
| 2007–08 | 10.04 | – | – |
| 2008–09 | 10.43 | – | – |
| 2009–10 | 9.54 | – | – |
| 2010–11 | 9.71 | – | 36,068 |
| 2011–12 | 9.59 | – | 33,740 |
| 2012–13 | 8.76 | – | 28,902 |
| 2013–14 | 7.55 | – | 28,337 |
| 2014–15 | 7.18 | – | 29,161 |
| 2015–16 | 6.53 | – | 32,724 |
| 2016–17 | 6.06 | 3.40 | 39,468 |
| 2017–18 | 6.25 | 3.25 | 47,467 |
| 2018–19 | 6.43 | 3.81 | 52,511 |
| 2019–20 | 5.66 | 3.68 | 55,170 |
| 2020–21 | – | – | 44,728 |
| 2021–22 | – | – | 49,189 |
| 2022–23 | 4.38 | 3.53 | 51,741 |
| 2023–24 | 4.59 | 3.18 | 53,995 |
| 2024–25 | 4.59 | 4.16 | 53,096 |
| 2025–26 | 4.33 | 4.48 | 48,774 |
Scroll the table sideways to see every column.
Annual data: Crime Survey incidents
Before 2002 labels are survey calendar years; from 2002 they are years ending March. Early surveys were not annual. Fraud starts in 2017; face-to-face survey gaps in 2021 and 2022 are left empty. Do not add these columns or substitute police records. A dash means no comparable observation is supplied; it is not zero. England and Wales, adults 16+ and households. These are incident estimates, not victim counts.
1996–2026; Million incidents; year labels explained below. No change is calculated across definition or assessment breaks.
| Year | Excluding fraud and computer misuse | Fraud |
|---|---|---|
| 1996 | No comparable observations supplied | |
| 1997 | 17.2 | — |
| 1998 | No comparable observations supplied | |
| 1999 | 15.2 | — |
| 2000–2001 | No comparable observations supplied | |
| 2002 | 12.8 | — |
| 2003 | 12.4 | — |
| 2004 | 11.6 | — |
| 2005 | 10.8 | — |
| 2006 | 10.9 | — |
| 2007 | 11.2 | — |
| 2008 | 10 | — |
| 2009 | 10.4 | — |
| 2010 | 9.5 | — |
| 2011 | 9.7 | — |
| 2012 | 9.6 | — |
| 2013 | 8.8 | — |
| 2014 | 7.6 | — |
| 2015 | 7.2 | — |
| 2016 | 6.5 | — |
| 2017 | 6.1 | 3.4 |
| 2018 | 6.2 | 3.3 |
| 2019 | 6.4 | 3.8 |
| 2020 | 5.7 | 3.7 |
| 2021–2022 | No comparable observations supplied | |
| 2023 | 4.4 | 3.5 |
| 2024 | 4.6 | 3.2 |
| 2025 | 4.6 | 4.2 |
| 2026 | 4.3 | 4.5 |
A scam can reach you without anyone entering your home.
Fraud is the largest category among the offences measured here. Repeat targeting matters: someone defrauded several times can appear several times in the incident count, but only once in the count of people affected.
Survey-based evidence
Crime Survey for England and WalesFieldwork: year to March 2026A household survey of adults aged 16 and over in England and Wales; estimates have sampling uncertainty.
How much of the crime the survey measures is fraud?
Fraud accounts for 47% of the offences measured here
47 in 100
headline survey incidents are fraud
Gold: fraud · Grey: other headline crime.
Rounded grid; exact share 46.6%. Incidents, not victims or a measure of harm.
Fraud compared with each other crime type
Millions of incidents · 2025–26 · same scale for every bar
- Fraud4.48m
- Computer misuse0.80m
- Criminal damage0.65m
- Robbery0.07m
4.48m fraud incidents ÷ 9.62m headline incidents. Component estimates are rounded.
Scope: England and Wales · adults aged 16+ living in households · Crime Survey headline categories only. Sexual offences, domestic abuse, stalking, harassment, crimes against businesses and homicide are measured separately and are not in this graphic.
Fraud makes up nearly half the incidents counted here. That is not half of all crime: the survey leaves out several serious harms and counts repeat incidents separately.
Show the workings — Fraud accounts for 47% of the offences measured here
Crime Survey for England and Wales headline incidents, the year to March 2026: fraud and the rest of the headline total. The grid shows the rounded share; the bars compare individual crime types.
Fraud is the largest headline survey category: 4.5m of 9.6m incidents. Its share has grown from 30% in 2016–17 partly because fraud rose 32% while most other crime fell. This counts incidents, not victims or harm: one person can be defrauded several times, and a lost card and a burglary each count once. It is not “half of all crime” — many harms sit outside the survey.
- Geography:
- England and Wales — adults aged 16 and over living in households (household crimes such as burglary are counted per household)
- Measure:
- Incidents (millions) and share of the headline total (%)
- Period:
- Interviews April 2025 to March 2026, about crimes in the 12 months before interview
- Status:
- Survey-based
- Calculation:
- Fraud incidents (4,483,000) ÷ all CSEW headline incidents including fraud and computer misuse (9,615,000) = 46.6%. ONS publishes both in thousands, so the share is “about” 47%.
- Compare with:
- 30% in 2016–17, the first year fraud was measured
- Not included / breaks:
- Not covered: crimes against businesses and public bodies, children under 16, people not living in households (for example in care homes or prisons), and homicide. Sexual offences, domestic abuse, stalking and harassment are measured separately and are not in this total. Police-recorded crime is a different measure and is not added in.
Source: ONS, Crime in England and Wales: Appendix tables, year ending March 2026, Table A1a (released 23 July 2026).
View the data as a table
| Part of the headline total | 2025–26 (millions) | Share |
|---|---|---|
| Fraud | 4.483 | 46.6% |
| Theft (including burglary and vehicle theft) | 2.628 | 27.3% |
| Violence with or without injury | 0.989 | 10.3% |
| Computer misuse | 0.798 | 8.3% |
| Criminal damage | 0.646 | 6.7% |
| Robbery | 0.071 | 0.7% |
| All CSEW headline crime | 9.615 | 100% |
National trends set the context. Your local picture depends on which offences are reported and recorded, and where they happen.
Explore crime in your area →
Local figures count police-recorded crime. Reporting and recording differ between areas, so compare with care.
Why local figures differ from the survey
Police figures miss unreported offences. The Hampshire pages use police.uk street-level data for September 2025 to August 2026. Over the year to March 2026, these counts were 1.8% below the ONS count for the same areas, mainly because offences without a mapped location are excluded.
That brings us to the practical question: which changes could prevent harm, help victims and make the response work better?
Why it is happening
Our analysis
Fraud often crosses borders and relies on banks, phone networks and online platforms. Serious violence is concentrated in particular places. Different harms need different responses.
Who is affected most
Behind the incidents are people and households.
Survey-based evidence
Crime Survey for England and WalesFieldwork: year to March 2026A household survey of adults aged 16 and over in England and Wales; estimates have sampling uncertainty.
Which crimes reach the most people?
Fraud reaches more adults than in 2016–17; most other measured crimes reach fewer
Earlier year → latest year. Changes are percentage points (pp), not relative percentages. Adults and households are different denominators; small movements may reflect sampling uncertainty.
2025–262016–17
- Fraud (adults)7.8%Was 6.0%
+1.8 pp - Other household theft (households)2.2%Was 2.0%
+0.2 pp - Criminal damage (households)1.9%Was 3.5%
-1.6 pp - Computer misuse (adults)1.4%Was 2.9%
-1.5 pp - Violence (adults)1.3%Was 1.7%
-0.4 pp - Domestic burglary (households)1.1%Was 2.1%
-1.0 pp - Theft from the person (adults)0.7%Was 0.8%
-0.1 pp - Robbery (adults)0.1%Was 0.2%
-0.1 pp
Fraud affects more adults than in 2016–17, while most other crimes measured here affect fewer. Serious violence is less common but its harm is concentrated in particular places and groups.
Show the workings — Fraud reaches more adults than in 2016–17; most other measured crimes reach fewer
Share of adults (personal crimes) or households (household crimes) who were victims at least once in the 12 months before interview, 2025–26 (filled dot) and 2016–17 (open dot).
Prevalence counts victims rather than incidents, so repeat victimisation does not inflate it. Fraud reaches the most people: 7.8% of adults, about 3.8m people, up from 6.0% in 2016–17. Most other crimes reach fewer people than they did then. Higher-harm violence is far rarer and concentrated by place and age, which a national survey average cannot show.
- Geography:
- England and Wales — adults aged 16+ in households, or households, as labelled
- Measure:
- Per cent victimised at least once (prevalence)
- Period:
- 2016–17 and 2025–26
- Status:
- Survey-based
- Calculation:
- As published by ONS (Table A3a). Personal and household rates have different denominators and cannot be added.
- Compare with:
- 2016–17, the first year fraud and computer misuse were measured
- Not included / breaks:
- Serious violence and homicide are too rare for a household survey to measure well; children under 16 are not included.
Source: ONS, Crime in England and Wales: Appendix tables, year ending March 2026, Table A3a (released 23 July 2026).
View the data as a table
| Crime (denominator) | 2025–26 | 2016–17 |
|---|---|---|
| Fraud (adults) | 7.8% | 6.0% |
| Other household theft (households) | 2.2% | 2.0% |
| Criminal damage (households) | 1.9% | 3.5% |
| Computer misuse (adults) | 1.4% | 2.9% |
| Violence (adults) | 1.3% | 1.7% |
| Domestic burglary (households) | 1.1% | 2.1% |
| Theft from the person (adults) | 0.7% | 0.8% |
| Robbery (adults) | 0.1% | 0.2% |
What could change
Proposals to consider. The bodies named below would need to agree to deliver them; responsibility is not a recorded commitment. Follow the links for costs and evidence.
Safer streets, fewer scams
Examine ways to reduce street crime and scams.
Who could deliver it: Police, banks, platforms and relevant regulators.
Main trade-off and delivery needs
Privacy and false positives; harms cannot be assumed minimal.
Choose scam pathway, bank/platform partners, lawful data sharing, comparator and compensation safeguards; separate policing action.
Cost: The extra cost is not yet known.
Costs, drawbacks and evidence →Test targeted violence prevention on harm and community trust
Test a targeted approach to preventing violence in one police area, measuring harm and community trust.
Who could deliver it: Police forces and local partners in England/Wales..
Main trade-off and delivery needs
Disproportionate searches, low trust and displacement.
Identify a real delivery gap and compare hotspot cohorts; preserve proportionality safeguards. National funding alone neither establishes nor refutes the additionality of a specific local improvement.
Cost: The extra cost is not yet known.
Costs, drawbacks and evidence →Manage court throughput with prisons and probation capacity
Assess changes to court capacity alongside the extra demand they could place on prisons and probation.
Who could deliver it: MoJ, judiciary, HMCTS, CPS and probation/prison services in England/Wales..
Main trade-off and delivery needs
Fair-trial safeguards, scarce legal staff and prison crowding.
Produce a costed court/prison/probation flow plan; do not infer that fewer jury trials is the answer. Legal-policy changes need their own appraisal; do not rank constitutional rights by this total.
Cost: The extra cost is not yet known.
Costs, drawbacks and evidence →
See all options and what is still unknown →
Some problems have no assessed solution yet: see coverage and gaps.
Domain SWOT
Recorded
Budget 2026 bears on one of these findings
Show the findings
Strengths · 2 findings
Crime other than fraud has fallen a long way: Crime Survey incidents excluding fraud and computer misuse were 78% lower in 2025–26 than in 1995, the survey’s highest year.
England and Wales · 1995 to 2025–26
Violence prevention has been evaluated: Home Office serious-violence funding (Violence Reduction Units with hot spot policing) was associated with about 550 fewer sharp-object assault admissions among under-25s (−8.5%) in funded areas.
20 VRU areas, England and Wales · April 2019 to December 2024
Weaknesses · 10 findings
About eight in ten adults say crime nationally has gone up, up from about six in ten ten years earlier, while Crime Survey crime excluding fraud fell about a third. A gap between belief and measurement can pull policy towards perceived rather than measured risks; worry about fraud and antisocial behaviour remains real and is measured separately.
England and Wales · Year ending March 2026
More officers do not automatically create a more effective response. The NAO found earlier productivity initiatives had not produced sustained change across forces, with inconsistent standards, patchy technology adoption and limited evidence of what works.
England and Wales · NAO audit, 3 November 2025
40.9% of adults experienced or witnessed local antisocial behaviour. This includes witnesses, not just direct victims. An April 2025 questionnaire-context change inflated part of the increase, so the full rise cannot be interpreted as behavioural change.
England and Wales · YE Mar 2026
The NAO’s July 2026 audit found gaps in monitoring, information and responses to breaches. Better supplier performance alone does not establish that the public is protected. Expansion needs reliable records, timely fitting and a staffed response to breaches. Tag counts are an input; monitoring coverage, response times and subsequent harm are the tests. The audit does not estimate the crime prevented by an additional tag.
England and Wales · 2026-07-10 · risk
The NAO reported that about half of people in prison had an identified drug problem in April 2025. Its February 2026 audit found that prison operations and health services were not consistently working together. Security, treatment and recovery need to work together. The audit could not establish an overall value-for-money estimate; the prevalence figure is not an estimate of treatment effectiveness. Track access to treatment, continuity after release and harms alongside capacity.
England and Wales · 2026-02-04 · risk
Domestic abuse and sexual violence need their own outcome tracking: they are outside the headline survey crime total. The NAO’s January 2025 audit found that the earlier cross-government strategies had not improved victim outcomes.
England and Wales · NAO audit, January 2025 · risk
Budget 2026 bears on this
Money for misogyny prevention and for delivering the Violence against Women and Girls Strategy; no outcome tracking is attached.
Fraud is the most common crime the survey measures: 4.5 million incidents and 3.8 million adult victims in 2025–26, 47% of all survey incidents — up from 3.7 million incidents in 2019–20.
England and Wales · 2019–20 and 2025–26
Only 8.5% of police-recorded crimes led to a charge or summons in 2025–26, against 15.5% in 2014–15.
England and Wales · 2014–15 and 2025–26
Cases wait longer to be heard: 80,829 criminal cases were open in the Crown Court at the end of June 2026, the most on record, against 38,341 in December 2019.
England and Wales · December 2019 and Jun 2026
Prisons are close to full: 86,779 people were in prison on 5 October 2026 against a useable capacity of 89,012 places, leaving 2,233 free — 97.5% full.
England and Wales · Monday 5 October 2026
Opportunities · 1 finding
Treat the two harms separately: stop scams where they pass through banks, telecoms and platforms, and focus patrols and problem-solving on the few places where violence clusters.
Depends on: Data sharing has a clear lawful basis and is judged on victim losses, and hot-spot work is chosen by harm, tested against comparison places and watched for effects on trust.
our synthesis
Threats · 2 findings
Scams adapt faster than controls: the 2023 Fraud Strategy aimed to cut fraud to 10% below 2019 levels by the end of that Parliament, yet survey fraud rose instead.
risk
Poorly targeted enforcement can move crime or damage trust: the main hot spots review did not assess community effects, and English results swung from a 7% fall in 2021–22 to no significant effect in 2022–23.
risk
Domain action plan
Proposed
Possible steps for government and the bodies responsible. These describe the order of work, not promised dates: starting an action and seeing results take different amounts of time.
Short term · prepare
Identify crime patterns and gaps in prevention and case handling.
Medium term · deliver
Test targeted prevention and improve the route from reporting to justice.
Long term · judge results
Track harm, repeat victimisation and whether people receive justice.
Delivery steps and owners (4)
- Prerequisite: Specify one scam pathway, its lawful data flows and an evaluation design · Home Office and the Online Crime Centre, with participating firms · proposed
- Decision: Select hot spots by harm, with comparison places and community oversight · Chief constables and Police and Crime Commissioners, with VRUs · proposed
- Delivery: Run the scam pilot; deploy patrols and problem-solving in short evaluated cycles · Participating firms and the Online Crime Centre; police forces and partners · proposed
- Result: Fewer fraud victims and losses; less serious violence in target and neighbouring places · ONS Crime Survey; police and NHS data · proposed
Sources, method and changes
Shareable visual summary
Visual summary
Recorded knife offences fell 6%. Fraud rose 11%. Do we feel safe?
England and Wales · Year ending June 2026, each compared with a year earlier: police-recorded knife or sharp instrument offences −6% (48,615, against 51,567); fraud incidents estimated by the Crime Survey for England and Wales +11% (4.5m, against 4.1m), a survey estimate. Police records and survey estimates are different measures and are not added together.
Changes compared with the previous year
- −6%police-recorded knife offences, year ending June 2026
- +11%Crime Survey fraud incidents (survey estimate), year ending June 2026
Each change is against the same 12 months a year earlier. Police records and survey estimates are different measures.
Text and data version
Recorded knife offences fell 6%. Fraud rose 11%. Do we feel safe? England and Wales · Year ending June 2026, each compared with a year earlier: police-recorded knife or sharp instrument offences −6% (48,615, against 51,567); fraud incidents estimated by the Crime Survey for England and Wales +11% (4.5m, against 4.1m), a survey estimate. Police records and survey estimates are different measures and are not added together.
| Measure | Value | Period | Geography | Type | Source |
|---|---|---|---|---|---|
| Police-recorded knife or sharp instrument offences | 48,615 | YE Jun 2026 | England and Wales | official estimate | Office for National Statistics, Crime in England and Wales: Appendix Worksheet B6 |
| Police-recorded knife or sharp instrument offences, a year earlier | 51,567 | YE Jun 2025 | England and Wales | official estimate | Office for National Statistics, Crime in England and Wales: Appendix Worksheet B6 |
| Change in police-recorded knife offences on a year earlierCalculation: (latest − a year earlier) ÷ a year earlier × 100 | −6% | YE Jun 2026 | England and Wales | derived calculation | Office for National Statistics, Crime in England and Wales: Appendix Worksheet B6 |
| Crime Survey fraud incidents | 4.5m | YE Jun 2026 | England and Wales | survey estimate | Office for National Statistics, Crime in England and Wales: Appendix Table A1a |
| Crime Survey fraud incidents, a year earlier | 4.1m | YE Jun 2025 | England and Wales | survey estimate | Office for National Statistics, Crime in England and Wales: Appendix Table A1a |
| Change in Crime Survey fraud incidents on a year earlierCalculation: (latest − a year earlier) ÷ a year earlier × 100 | +11% | YE Jun 2026 | England and Wales | derived calculation | Office for National Statistics, Crime in England and Wales: Appendix Table A1a |
Available images and their source records are listed on Shareable visuals.
Sources and methods (4)
- ONS, Crime in England and Wales: year ending March 2026 (released 23 July 2026) — headline totals, statistical significance, scope and the reporting rate
- ONS, Crime in England and Wales: Appendix tables, year ending March 2026, Tables A1a, A3a and A5a; Worksheet B6 (released 23 July 2026) — Crime Survey incidents by type (A1a) and share victimised (A3a); police-recorded homicide (A5a) and knife or sharp instrument offences (B6)
- Home Office, Crime outcomes in England and Wales, year ending March 2026, Table 1.3 (released 23 July 2026) — charge rate and other outcomes
- Ministry of Justice, Criminal court statistics quarterly: April to June 2026, Table C1 (released 24 September 2026) — Crown Court open caseload
Method notes
- The Crime Survey for England and Wales asks people aged 16 and over in households about crimes they experienced in the previous 12 months, reported or not. It is the best guide to long-run trends in the crimes it covers, but it misses crimes against businesses, children and people not in households, and is too small to measure rare crimes such as homicide.
- Police-recorded crime counts what is reported and recorded. It covers more offences and is the only source for local figures and rare, serious crimes, but it moves with reporting and recording practice as well as with crime.
- Survey and police figures are never added together or put on one axis on this site.
- Police-recorded fraud and computer misuse are temporarily not published while the City of London Police moves to the new Report Fraud system (ONS, July 2026).