Thank you for this excellent analysis. The work you have done should be read by all of those people including intelligent friends of mine, who think that London is a very dangerous place.
I am grateful for your voice of sanity in a world of exaggeration and lies.
Thanks for your kind words. It's amazing how this deranged "London has fallen" narrative is believed by so many people, in defiance of all the evidence. IMO it is the greatest city that there has ever been; now more than ever.
Murders are overwhelmingly committed by people known to the victim. You're barking, to a certain extent, up the wrong tree. A better metric would be *assaults* over, say, a decade (long term trend = upwards). Then you would need to examine the way game stats (via Home Office Counting Rules and National Crime Recording Stats).
As it stands, your optimism regarding 'safety' strikes me as Panglossian and your faith in statistics naive.
Typically interesting article. It would be interesting also to know if the types of murder/homicide have changed over the years. For example, the percentage of cases where the perpetrator was known to the victim. And the percentage unsolved. I suspect "the streets" are even safer - at least from a gory death - than Fraser suggests, as it seems that many homicides are carried out after arguments with someone known to the victim - for women, sadly, often their partner or ex. Killings by a stranger are probably even rarer than ever.
Great to see more data on this, I've been looking into homicide/murder figures globally and in the last few years it seems to be trending down in the majority of countries (that have recent data).
Country----------2024 Rate(per 100k) --2025 Rate so far--(Year on Year change)
Thank you Fraser, another excellent factual analysis that encourages critical thinking about the political narratives of some on the right. As an almost life-long Londoner I would say that physical violence and murder are in decline. I had forgotten the surge around 2018, but the trend is certainly downward. Even lesser offences (car theft was the big one for the average Londoner) have dropped off.
I think the change in the average age of Londoners is important - and our comparative prosperity. In comparison to the most deprived areas of the UK, (where I have family), London is an enjoyable place to grow up and offers many more legal opportunities to bored young men, who might otherwise be drawn into crime.
And the Survey data was interesting too! London residents reporting half as much crime now as in 2000!! Imagine turning that chart into a story of London being "finished"
Great piece - the fear of crime has certainly taken hold and there's probably something to be said for more street crime - moped/phones snatched etc... but the numbers are good news
Fraser, I applaud you (and Robert Colville and others) for still trying to get through to people on Twitter, though comparing the comments there to here is, well, radically (and the correct word to choose in more than one way) different!
Thanks - a really interesting article - and it ties in with my experience of being in London for 25 years or so.
If you want another interesting theory to look at - have a look at the impact of lead on crime - https://freakonomics.com/2007/07/lead-and-crime and https://www.bbc.com/news/magazine-27067615 - I'm not sure how much the reduced lead in London after the end of of lead paint helped the crime reduction, but it's interesting and ties in with the data you've got here.
I think your underestimating the amount of work to compile that data. Likely more a masters or PhD, if you want to look at the why as well, than a School Project.
This was one of the original aims of the murdermap.co.uk website - but the work involved solely to keep track of new cases meant that the attempt to build a comprehensive historical database has not yet been realised
A non-academic study of homicides in London in 1888 (28 murders, 94 manslaughter according to the Met commissioner at the time) suggested a significant number would be treated today as death by dangerous driving or infanticide. Only around half were publicly reported (newspaper / court records) and as Fraser notes the archive register of cases only begins a year or three later
Thank you for this excellent analysis. The work you have done should be read by all of those people including intelligent friends of mine, who think that London is a very dangerous place.
I am grateful for your voice of sanity in a world of exaggeration and lies.
Thanks for your kind words. It's amazing how this deranged "London has fallen" narrative is believed by so many people, in defiance of all the evidence. IMO it is the greatest city that there has ever been; now more than ever.
Seconded...
Agreed!
Murders are overwhelmingly committed by people known to the victim. You're barking, to a certain extent, up the wrong tree. A better metric would be *assaults* over, say, a decade (long term trend = upwards). Then you would need to examine the way game stats (via Home Office Counting Rules and National Crime Recording Stats).
As it stands, your optimism regarding 'safety' strikes me as Panglossian and your faith in statistics naive.
* the way police game stats
Typically interesting article. It would be interesting also to know if the types of murder/homicide have changed over the years. For example, the percentage of cases where the perpetrator was known to the victim. And the percentage unsolved. I suspect "the streets" are even safer - at least from a gory death - than Fraser suggests, as it seems that many homicides are carried out after arguments with someone known to the victim - for women, sadly, often their partner or ex. Killings by a stranger are probably even rarer than ever.
Great to see more data on this, I've been looking into homicide/murder figures globally and in the last few years it seems to be trending down in the majority of countries (that have recent data).
Country----------2024 Rate(per 100k) --2025 Rate so far--(Year on Year change)
Jamaica-------------40.1-------------------23.3-------------(−42%)
Canada-------------1.9--------------------1.4---------------(−23%)
United States-------5----------------------4-----------------(−20%)
Honduras-----------25.3------------------20.5--------------(−19%)
The Philippines-----4.3--------------------3.5---------------(−19%)
Mexico--------------19.3------------------16.2--------------(−16%)
South Africa--------43.5-------------------38.1--------------(−12%)
Spain----------------0.61------------------0.54--------------(−12%)
Bangladesh---------2.4--------------------2.2---------------(−8%)
England and Wales-0.94------------------0.88--------------(−6%)
Costa Rica----------16.6-------------------16---------------(−4%)
Ecuador-------------38.8------------------57----------------(+47%)
More data/sources here - https://gdea.substack.com/p/global-homicide-2025-so-far
Thank you Fraser, another excellent factual analysis that encourages critical thinking about the political narratives of some on the right. As an almost life-long Londoner I would say that physical violence and murder are in decline. I had forgotten the surge around 2018, but the trend is certainly downward. Even lesser offences (car theft was the big one for the average Londoner) have dropped off.
I think the change in the average age of Londoners is important - and our comparative prosperity. In comparison to the most deprived areas of the UK, (where I have family), London is an enjoyable place to grow up and offers many more legal opportunities to bored young men, who might otherwise be drawn into crime.
Maybe I’ll risk saying the unsaid….has someone in the Met found a cunning way of excluding ‘black on black’ stabbings from the stats?
Hi Alan, the figures I show is not from the police - its NHS data for those treated in hospital for knife attacks
Ah I see. Thanks.
And the Survey data was interesting too! London residents reporting half as much crime now as in 2000!! Imagine turning that chart into a story of London being "finished"
Great piece - the fear of crime has certainly taken hold and there's probably something to be said for more street crime - moped/phones snatched etc... but the numbers are good news
Hi Fraser. What happened to your article on the lifetime value of benefits?
Fraser, I applaud you (and Robert Colville and others) for still trying to get through to people on Twitter, though comparing the comments there to here is, well, radically (and the correct word to choose in more than one way) different!
Good piece, Fraser. This may be of interest: https://thecritic.co.uk/crime-rates-really-have-fallen/
Thanks - a really interesting article - and it ties in with my experience of being in London for 25 years or so.
If you want another interesting theory to look at - have a look at the impact of lead on crime - https://freakonomics.com/2007/07/lead-and-crime and https://www.bbc.com/news/magazine-27067615 - I'm not sure how much the reduced lead in London after the end of of lead paint helped the crime reduction, but it's interesting and ties in with the data you've got here.
I think your underestimating the amount of work to compile that data. Likely more a masters or PhD, if you want to look at the why as well, than a School Project.
Perhaps I'm swayed by my son spending an entire summer on his school project! (Which was excellent btw)
This was one of the original aims of the murdermap.co.uk website - but the work involved solely to keep track of new cases meant that the attempt to build a comprehensive historical database has not yet been realised
An example of an imperfect attempt to chart London murder / homicide stats is here https://www.murdermap.co.uk/statistics/london-murder-statistics-historical/
A non-academic study of homicides in London in 1888 (28 murders, 94 manslaughter according to the Met commissioner at the time) suggested a significant number would be treated today as death by dangerous driving or infanticide. Only around half were publicly reported (newspaper / court records) and as Fraser notes the archive register of cases only begins a year or three later
What are the wider trends for each of England, Scotland, Wales and Northern Ireland?
London's uniquely different from the larger UK.
England and Wales (including London) have a similar trend - https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/bulletins/crimeinenglandandwales/yearendingmarch2025#homicide
NI Has the distorting impact of the troubles
With a population MUCH Smaller than London it saw about 100 deaths a year for approx the second half of the troubles.