Is violent crime rising?
The best long-term measure shows violent crime has fallen by more than three quarters since the 1990s. Most people think it is going up.
What has happened to violent crime in England and Wales over thirty years, on which measure, and why does the direction of travel feel like the opposite?
Part 1
What this is about
There are two main ways of counting crime in England and Wales. One is the police’s own records. The other is a big survey that asks households what has happened to them.
Police figures only include crimes people report, and they change when police change how they record things. The survey also picks up crimes that never get reported, so it is better for seeing long-term trends.
On the survey, violent crime peaked in 1995 at about 4.5 million incidents. By the year to March 2026 it was 989,000. That is a fall of 78%.
Yet in 2013, after most of that fall had already happened, 51% of people thought violent crime was rising.
4.5m
Violent incidents at the 1995 peak
CSEW
989k
Violent incidents, year ending March 2026
CSEW
−78%
Change from peak to latest
CSEW
Explainers
Explainer video
Many people think violent crime is rising. The long-term figures say the opposite.
Summary infographic
Is violent crime going up?
A big survey of households is the best way to see long-term trends in violent crime.
4.5m — Violent incidents at the 1995 peak — CSEW
989k — Violent incidents, year to March 2026 — CSEW
−78% — Change from peak to latest — CSEW
51% — Thought violent crime was rising (2013) — Ipsos
The trick: Picking the measure and the years
Police figures can rise because more crime is reported or recorded, not because more crime happens. A short run of years can point the opposite way to the long-term trend.
Violent crime is out of control and rising.
Violence has fallen by more than three quarters since 1995.
It fell a further 34% between 2016 and 2026.
In 2013, 51% of people thought it was rising.
Violent crime has fallen by more than three quarters since 1995. Watch for claims that pick a different measure or a short run of years.
Part 2
What's being said
51% of the public believed violent crime was rising.
Ipsos for the Royal Statistical Society and King’s College London, 2013
By 2013 the CSEW measure had already fallen by around two thirds from its peak.
Asked what share of violent deaths are caused by a knife, the average answer was 71%. The recorded share was 25%.
Ipsos, Perils of Perception 2018
Asked how full prisons in England and Wales were, the public guessed 143% of capacity. The figure was 108%.
Ipsos, Perils of Perception 2018
“Violent crime is out of control and rising.”
Paraphrase of a framing in wide circulation
Source neededHow it was used
Used to argue that the country is becoming more dangerous.
How it could mislead
It usually leans on police-recorded figures or a short run of years. Those can rise because more crime is reported or recorded, while the long-term survey shows a fall.
What the figures say
On the survey built to track long-term trends, violence is down 78% from its peak, and down another third since 2016. Saying it is rising means choosing a different measure or a short window.
“Most violent deaths in this country involve a knife.”
Paraphrase of a framing in wide circulation
Source neededHow it was used
Used when knife crime is in the news, to suggest knives are behind most killings.
How it could mislead
Knives are the most common method, which makes the claim sound right. But “most” is nearly three times the real share.
What the figures say
About a quarter of violent deaths involve a knife, which makes it the most common method. But “most” is nearly three times the real share.
Part 3
What the figures show
A big fall since the 1990s
The chart only shows years with a published survey figure. We don’t draw a line between them, because the survey’s definitions have changed over time.
Even so, the direction is clear. Most of the fall came before 2016, and it has kept falling since, by a further 34%.
CSEW, violence with or without injury · sourced anchor years only
Sourced anchor years only — no intermediate values are shown.
What people believe
In 2013, 51% of people thought violent crime was going up, even though it had already fallen a lot.
With knife crime, people get the size wrong. They guess knives cause 71% of violent deaths. The real figure is 25%.
Ipsos, Perils of Perception 2018
Knife deaths as a share of violent deaths
×2.8
Prison population as a share of capacity
+35 pts
The other side
- The survey doesn’t count everything It only covers people living in households. It leaves out murders, crimes against businesses and some specific offences.
- A national fall isn’t a fall everywhere Some areas, age groups or types of crime may not have improved. Someone describing their own street isn’t contradicted by a national figure.
- The belief figure is old The 51% figure is from 2013. People’s views may have changed since.