Is Britain becoming more violent?
A look at the data
After atrocities like the Huntingdon train stabbing there is shock - and attempts to make sense of what happened. Perhaps demands for ministers to pass a new law, or for police to change tactics. But what conclusions can be meaningfully drawn at this stage? What trends, if any, are discernible?
The police, now, are quick to state the ethnicity and nationality of the suspect held: black and British. Matt Goodwin, a GB News presenter, puts “British” in inverted commas. His argument is that many recent killers, even if born here, are of non-British descent so their crimes are indirect result of immigration. This is a common way of seeing crime in Sweden where the phrase ‘foreign background’ is used to bracket together immigrants and their offspring.
But while Swedish gun crime has soared, British violent crime - knife, gun, common assault - has been falling. High-profile cases shared on social media can give the opposite impression and perspective is easily lost.
In a post entitled ‘when will it end?’Goodwin says he learned about the train attack, from….
“…a text from a friend who lives outside the United Kingdom. “What on earth is happening to your country?” I turned on the news. Another mass stabbing, this time on a London-bound train, which put ten people in hospital, with life-threatening injuries. Three words, I replied. Mass uncontrolled immigration. That’s what happened to our country.”
Social media allows such theories to be shared without any fear of contact with actual facts. But for those interested, here are the facts about recent trends in violent crime.
First of all, the immigrant population has certainly doubled over the last two decades. But crime, as measured by the Crime Survey for England and Wales (CSEW), has halved.
This collapse in crime has come in a country with more children then ever born to immigrants: about a third of all births now, rising two-thirds in London. The 30,000-household Crime Survey is gold-standard because it’s consistent - unlike police-recorded crime, where definitions change and it’s not designed to be comparable year-on-year. But a survey is only a sample. The CSEW has struggled with sample sizes post-pandemic, jars with police-recorded crime (which shows many crimes soaring) and has other flaws documented by academics (see this Lancaster paper).
Nigel Farage dismisses the CSEW on this basis and says - in effect - look at your phone! Crime is soaring! So given the flaws in CSEW, who do we believe?
Enter the NHS. Its has hospital data on who is treated for various forms of violence (assault, knife attack, bodily force, firearms etc). I pulled the below from the NHS database. Like the CSEW it shows violence trending down - for about two decades. Click on the menu on the below chart to see different NHS data for different forms of violent crime.
After Huntington, the Conservatives say that knife crime is too high. I agree. But the latest figures show hospital admissions for knife crime injuries at a 25-year low. London, the most immigrant-heavy and racially diverse city in the UK, releases monthly figures for murders (perhaps the hardest of all crime metrics). They are at a multi-year, perhaps multi-century low.
And what about perception? Conspicuous low-level crimes (like shoplifting, tailgating) are rising, after being unwisely deprioritised. This projects a sense of lawlessness: seeing a leg of lamb security-tagged in Tesco does not inspire confidence. You don’t see a drop in assault or knife crime, but social media means we certainly do see videos of crimes a lot more easily.
If Matt Goodwin’s overseas friend gets his news from Twitter he may well feel that the UK is gripped by a wave of migrant-driven violent crime. So how to see the context? We could consult large-scale surveys asking people whether they feel safe in their local streets. It shows general perception of improvement over three decades, especially from women.
We are still trying to find out what happened on that train and understand what motivated the train killer. But it’s hard to find any evidence supporting the idea of a national crimewave. Scholars at Cardiff University’s Violence Research Group have looked at this from every angle for their 2025 report (pdf) and concluded:-
“Over the past two and a half decades … serious violence in England and Wales has decreased substantially. This message needs to be much better known, not least because it reflects better prevention and because fear of violence, often stoked by reports of rare tragic violent events, corrodes individual and community wellbeing”.
Perhaps that’s what Goodwin could have said in reply to his friend. What’s happening to Britain? In the case of the Huntingdon train attack: an act of random, senseless, grotesque violence. The sort no country is ever able to inoculate itself against.
Goodwin could add that the death toll from that train was curtailed because an unarmed guard, Samir Zitouni, confronted the knifeman and now lies critically ill in hospital. Yes, immigration has brought us problems. But, also, heroes.
The overall story in Britain - as far as two decades of police, survey and hospital records can attest - is of violent crime in steady decline. And strange as it may sound to say it, the streets of these islands are perhaps safer now than they had ever been.
PS For those interested, a full list of crime data with links back to official sources can be found in the data library page on this website.


The NCS certainly was the gold standard. But perhaps it is fools gold. The response rate has fallen from about 75% pre-pandemic to around 42% in 2023. It does not survey those in communal accommodation (the young, and many recent immigrants). By focusing on households (even among which, the response rate has dropped precipitously), it fails to capture many of those that are arguably most likely to be the victims of crime.
There are good reasons to think that Britain is safer today for the average middle class Brit than it was in the past. But that does not mean that it is for everyone. In particular, ironically for recent immigrants. And it is not at all clear that the impact of crime on them (and committed by them on others in the same socio economic groups) are being recorded by the NCS.
Since COVID, screens have remained separating bus drivers from the public. We see reports of increases in attacks on staff in the NHS and rail. Anecdotally at least, those in non-police rule enforcement roles are increasingly reluctant to enforce those rules due to fear of violence on the part of those they are enforced against. It seems hard to claim that everything is getting better when those in such roles feel that it is not.
I wonder if the methods of statistical data gathering that worked in the past are still functioning in the way they should in that they increasingly only capture one part of the picture.
I'm afraid that the "gold-standard Crime Survey" isn't necessarily accurate. It relies on victims making the effort to report their victimhood to the police, and if they don't, perhaps because they have little confidence that police will actually do something - they won't bother. I certainly wouldn't and I was a London cop for many years.