This post aims to provide some historical context to homicide statistics across England and Wales, based on publicly available data. It will hopefully be updated every year.
Firstly, when we talk about homicides we count not only murder but also manslaughter and infanticide. A significant proportion of cases are investigated as ‘murders’ but end up as manslaughter cases because of a lack of intent to cause serious harm, the suspect’s diminished responsibility due to mental illness at the time or some other legal defence such as loss of control.
Also we have to recognise that the recording of statistics varies in consistency over the years for many different reasons (for example, are we are better at detecting homicides now than 100 years ago?) Having said that, homicides are perhaps the crime statistic least open to manipulation by police or other agencies. It’s therefore worth having a look how homicide has changed over time as it might help us to work out why it varies and how we can go about driving the number of homicides down towards the ideal level of zero.
The available statistics for homicide in England and Wales from 1960 onwards show the number of recorded offences peaking in 2003 before decreasing to a low point in the 2014, followed by an another peak from 2016 onwards which appears to have been halted by the Covid pandemic and may now be in reverse. Note that there was a change from counting by calendar year to financial year in the late 1990s, presumably so it matched the annual budget period for police etc.
That graph reveals how the number of homicides for a specific year itself can also change over time, as the cases are fully investigated – with a small proportion being recategorised as “not a homicide” (Blue shows the initial count of homicides at the end of the year and red shows the number of offences “currently recorded as homicide). It also shows the effect of major events on statistics gathered by the government, as it includes terror attacks and other types of mass killing.
These are:
- 15 victims of Michael Ryan (1987)
- 58 people who suffocated in a lorry on the way to the UK (2000/01)
- 172 victims attributed to Harold Shipman (2002/03)
- 20 cockle pickers who drowned in Morecambe Bay (2003/04)
- 52 victims of the London July 7 terror attacks (2005/06)
- 12 victims of Derrick Bird (2011/12)
- 96 victims of Hillbsorough (inquest verdict in 2016/17) and 4 in Westminster Bridge terror attack
- 31 victims of Manchester and London Bridge terror attacks and 11 victims of Shoreham air crash
- 39 human trafficking victims found dead in a lorry in Essex (2019/20)
The yellow line shows what the totals would have been without these incidents included.
If you examine the 100 years between 1898 and 1997 you can see how homicide remained pretty steady (apart from spikes in 1942 and 1945) until the 1960s, when it shot upwards.
The population of England and Wales has grown steadily over the last two hundred years (there was no census data for 1941).
If we look at homicides per million population since 1970, to try and take account of this growth, we can see that there is still a rise to around 2003 before a drop (and then recent rise). As one commenter below has suggested, it would be interesting to see whether changes in specific age groups are more relevant given that there is a high proportion of murder victims (and suspects) aged between 16 and 30. This also applies to crime in general and is known to academics as the age crime curve – as in the number of crimes committed peaks at between 16 and 20 before decreasing through adulthood. If you look at the statistics page on the 2022 murder map you can see this in the age range graph of homicide victims.
So what caused that rapid growth in homicides from the 1960s onwards? Was it population change, the state of the economy, the new ‘permissive society’, a breakdown of ‘family values’, or the effect of hard drugs like heroin and cocaine? Or a complicated combination of these factors (and others, such as the way the data is recorded)? And why did that trend reverse from 2003? Heavier sentences? Better policing? Improvements in emergency medicine?
It has been suggested that the downturn since 2003 is down to a sustained fall in the level of domestic violence, given that a significant proportion of murders are carried out by partners, former partners or family members. (For the year 2021/22 the number of victims in England and Wales who knew the suspect was 301 out of 696 (of which 78 were partner or ex-partner and 71 were family of some kind). Perhaps there now seems to be less tolerance of violence in society.
Another possible factor is the improvement in the level and quality of “security” in society. Although this may apply more to other crimes which decreased from the 1990s onwards, such as burglary and theft (think car immobilisers, alarms, CCTV, phone locks etc), you can see how improvements in technology also help to detect and deter murder. Police can now link suspects to the scene of the crime by mobile phone location data and CCTV images as well as by witness statements and forensic evidence (which has also improved).
This post was originally published in 2012 (as you can see from the comments below). It was updated in March 2023 with recent statistics and will be updated every year. Future updates will hopefully involve charts for age ranges, weapons, relationships etc for recent years.
We have also looked at the historical data for London only, although at the moment it is only readily available from 1990.
Statistics from the ONS, Homicide in England and Wales, and Population estimates.
Looking at The Guardian’s article about the 80 women killed in the last year, I was led to wonder about the cluster of elderly women. Is there an uptick in killings of wives by elderly husbands, possibly linked to the increase in dementia and limits on support services/care etc?
If so, this has implications for political decision-making.
It’s a mix of husbands/partners and offspring (usually their son). Many of the offspring are found guilty with diminished responsibility.
The abolitionof the death penalty in the UK was put forward because people are civilised. People living in society are civilised but there are those amongst us who when commiting vile crimes do not deserve to live. A rapist who slaughters his victim, child killers, killing for money. Do we keep these people alive merely because we are civilised? There comes a time within society when enough is enough and a deterrent and a consequence to actions must the right thing. The victims are being failed in society now in our vain attempt to prove we are ‘civilised’ there are those who never will be civilised and stalk their victims knowing the ultimate deterrent is not there.
The official abolition of the death sentence in Britain was in 1965. However, the functional abolition of the death sentence (meaning the death sentence no longer applied by default to intentional murder, and therefore was not a deterrent to homicide in the true sense of the term) was in 1957. That is exactly the inflection point year on the graph.
I am from the States, and even in this country the Supreme Court, whilst allowing capital punishment in extreme cases, has banned mandatory death sentences for first degree murder or any other crime as being unconstitutional. Even notorious Texas only executes contract killers, cop killers, judge killers, rapist-murderers, torture-murderers, and such like. Straight first degree murder alone can only get you a life sentence, if that. As such, the only countries where a true test of the efficacy of the death sentence as a deterrent to homicide might be conducted are in East and Southeast Asian countries, where the deterrent seems to work. As for South Africa, it should be noted that the apartheid government became seriously squeamish about executing people after the 1960s, and the end of apartheid meant the end of low-grade guerrilla war in that country. Both of these will have affected their homicide rates.
On another note, the lead-homicide hypothesis broke down when data for the last twenty years is included, and the correlation ceases.
I will correct my previous statement about South Africa. The country went squeamish in the ‘30s, then really stepped up executions in the ‘80s, likely in response to the low-level insurgency then underway against white rule.
The answer is clearly lead neurotoxicity.
https://ricknevin.com/update-juvenile-crime-still-falling-fast-in-2021/
Lead damages the prefrontal cortex which is involved in executive functioning. Now that lead has been removed from the ambient air, crime rates, teenage fertility rates etc. etc. etc. have all been plunging to one all time low after another. The UK removed lead a few years later so its crime reduction happened a few years later.
Surprisingly, the 21st Century could become a time without youth crime.
The lead hypothesis has always fascinated me. Because I honestly can’t tell if it’s a great hypothesis, or just total bunkem. It’s true that there is a decent statistical correlation between lead and homicide rate in any given country [decreasing lead exposure resulting in lower homicide rates], however this could be a statistical artefact. I’m pretty sure if we took measurements of any number of contaminants, both industrial and not, that have decreased over the years, we would also see this type of conjoinment. Does this speak to a genuine relationship?
It seems odd that this PARTICULAR contaminant would be the focus, and not something else like asbestos or something more culturally and socially linked, for example gratuitous and gritty violence in movies. The 70s were a peak for that. Everything in violent media after that became very glossy and serialized [especially the advent of video games, despite their reputation]. Why lead in particular, as opposed to other neurotoxic chemicals? What about homicide rates prior to the patronage of lead for industrial uses? Homicide rates appeared to be extremely high per capita in Middle Ages Europe, where I don’t think lead is a great candidate for the explanation.
The link between lead in the air and the murder rate is well established, You have to remember lead was widely distributed in the air and virtually everybody would be breathing it in unlike other pollutants. As for the 70s being a peak for violent media I really couldnt agree, take a look at programs in the 70s , even when people were shot there was rarely even any blood shown let alone anything more realistic, the violence available at the click of a button now is far, far worse.
@Steve Ryan “Or why it is that countries which have abolished the death penalty in the last 20 years (eg Albania, Bulgaria, Latvia, Poland, South Africa, Ukraine, etc ) have all seen a decline in their homicide rates since then?”
I can’t comment on South Africa, but the other countries you listed are all post-Soviet countries which were (and to some extent still are) dealing with very high levels of poverty, political instability and corruption. As the situation has improved, it’s hardly suprising that the murder rate has dropped, regardless of what else happened during that time. Most of those countries, Poland aside, still have pretty high murder rates.
In response to Jack and Richard, these figures are for the homicide rate, which includes manslaughter as well as murder.
Whenever the statistics for homicide start rising the liberal left of the UK who now control our justice system change the definition of murder they ineffect move the goal posts so what was murder yesterday is manslaughter or a lesser charge today, when the truth of the matter in real terms is that homicide is in a perpetual upward spiral.
I doubt that as a general rule. I think times that are more grungy, gritty and dank tend to have a higher homicide per capita rate.
London was a lot more grotty in the 1980s, and had a higher per capita homicide rate than the 2000s. The homicide rates in NYC have decreased precipitously from the 1970s-80s period through to the early 2000s. God bless modern social media [YouTube and the like] for being able to glimpse at the footage of the NY subways during the 1980s, EXACTLY as I remember them during my visits there. Dark, low lighting, graffiti everywhere, the smell of urine [not represented], menacing vibe. The 30 second intro of the old Equalizer series captured this vibe perfectly. And the same is true of London. When I look at the Tube today, with it’s spotless seats [have you SEEN the Elizabeth line?! You could practically eat off the floor. So much space that the entire carriage could be packed full of Big Daddy wrestler types and you’d still have room left over] and furnished interior decor… I think back to the Old Tube. Dimly lit, lights go out passing through a tunnel, horrible grimy smell and seats. Uncomfortable. “Handlers” passing through carriages and starting trouble with people, no cameras or police around. The rattling in the carriages at high speed, trains feel like they are about to de-couple. Much more dangerous times. I think people just have nostalgia goggles when it comes to this stuff.
I know the discussion was in 2018 but what a balanced intelligent discussion.
It would be useful in this debate to look at the figures for manslaughter as this plea is used by defence lawyers to avoid a murder charge and a mandatory life sentence. It is likely that many of the murder cases in days gone by would now be downgraded to manslaughter. This of course will skew the figures but the result is the same. Namely that the victims have lost their lives.
Perhaps those above who blame the increase in homicide rates on the abolition of the death penalty in 1965 can explain why, in the USA, those states that have kept the death penalty have higher homicide rates than the states that have abolished it? Or why it is that countries which have abolished the death penalty in the last 20 years (eg Albania, Bulgaria, Latvia, Poland, South Africa, Ukraine, etc ) have all seen a decline in their homicide rates since then?
I mean be careful about looking at it like that. Chicken-and-egg problem. Was the death penalty kept because the rates of homicide were so high already? Southern states have laxer [generally] gun laws and more death penalty [broadly generalizing], but I wouldn’t put their generally higher homicide rates down to either of those things.
You really have to look at a place with the same or similar demographic profile [age, sex, race, ethnicity, poverty level] and then analyze it over time between death penalty/abolition.
Global warming?
The vast majority of murders these days in London and other big cities are young black men or teenagers killing other young black men or teenagers. Homicide involving other categories of people are extremely rare, almost vanishingly so given the huge populations of people we’re talking about.
Don’t think you can use the words “vast majority”. Take a look at the list put together by the BBC (https://www.bbc.co.uk/news/uk-46128268) and you can see for yourself.
Of the 126 victims we have counted, around a quarter of them are women, which is still pretty significant.
Of the 98 men, just over half (51) are aged between 11 and 30. (Roughly a third of all victims are aged between 16 and 24).
It’s been widely reported that there has been a significant increase in domestic murders this year (although last year was a record low).
But yes homicide in general is very rare in London when looked at as a ‘rate’ per population.
I think Andrew means that amongst teenagers, homicide is likely to be largely young black males, often involved in gang disputes and turf wars. I think that’s basically a solid point, if you collect a mugshot of young offender homicides in London.
As for the overall, they are likely to be disproportionately represented. That would be a fairly accurate assessment too.
But homicide in London is actually pretty rare per capita, given the size and demographic profile of the City of London. In contrast to popular depictions of it.
It really seems to me that at least 300 innocent people are killed each year due to our determination to spare the lives of those prepared to kill. I cant see how this can be justified
I agree with you wholeheartedly Brian except that the figure for 1960 homicides was about 300. The current figure is well over 600 and peaked at about 700 in 2020
The population increased by around 30% so it is not like for like
1915 USA less than 300 homicide deaths 2015 14000 ? Yet medical advances make this US figure even more horrific. Crimes in Australia show an increase by criminals using knives. With the explosion of ICE drug usage robberies of small amounts of money has escalated. In WA 2018 there has been a 50 % decrease in ICE taking from 2017 due to severe penalties and education .
Don’t forget advances in medical technology and incident response. IMO these are a major factor – probably THE major factor – in the recent fall in homicides. Violence itself is on the increase, despite the desperate attempts of the authorities to cover it up, especially in the big cities. If you compare medical care now with medical care 50 years ago, the discrepancy is so great that the homicide figures cannot even be compared.
Surely, two of the biggest factors in the rise in the murder rate post 1960 have been 1) criminalizing drug use and drug dealing in 1960 (or thereabouts), and 2) the steady rise in mass immigration to this country? The latter not accompanied by any significant increase in checks on exactly who is coming in. I’m not even sure the abolition of hanging has been so significant, even though I believe it should be brought back for serial killers and terrorists. It seems to me that we turned the UK into a huge market for illegal drugs back in 1960 and thereby attracted criminals from around the world, and then made it relatively easy for them to come here with their contraband. Whilst many on the Left in politics want to blame inequality, poverty, government austerity etc., none of their reasons for the much higher murder rate really stack up. In the 1930s, for example, poverty was real not relative and social services nonexistent, yet the murder rate (even on a per capita basis) was less than half the current level.
Wouldn’t really explain the drop from 2003 onwards per capita though… since immigration boomed in the years 2003-2015 and yet homicide rates fell per capita during same period.
Do we have solid evidence for the effects of de-criminalizing drug use in lowering crime rates, especially homicide? I mean not simply studies analyzing prohibition-era policies, which tend to be sketchy investigations at best. But solid statistical evidence backed up by plausible hypotheses about the reasons for the effects?
I’ve always been extremely skeptical of the decriminalizing drugs focus groups, as they never seem to take counter-evidence seriously. Their stance is extremely uni-directional.
gun ban 1995. followed by a fifty percent increase in total homicide until 2004. (with less than 10% population increase for same time period)
ROFL. They take away rights, and crime explodes. Gun restrictions are retarded. AUS saw their rate of armed robberies triple within three years of their gun ban.
Homicide in London increased from the 1960s onwards, peaked in 2003, then declined by more than 50 per cent until 2014, and is now increasing again.
Handguns have never been used for self defence in the UK, you were never permitted to carry a handgun in public nor in your shop nor house, so your assumption is baseless.
Yh, that’s a good point. When talking about firearm bans we have to elucidate what the actual laws and culture of use is, because as you say, if handguns are not used/allowed as self-defence weapons it’s kind of a moot point when discussing firearm bans [say you’re allowed for things like hunting or recreational sport].
Plus, didn’t Australian mass shooting drop precipitously after their firearm ban? Also, what were the rules on use for self-defence in Aus [anything like the US]? Was it common for regular citizenry in Australia to defend themselves using firearms?
This is incorrect. There were no serious restrictions on owning firearms in the UK until after the First World War. To quote Blackstone (courtesy Wikipedia), ‘The fifth and last auxiliary right of the subject, that I shall at present mention, is that of having arms for their defence, suitable to their condition and degree, and such as are allowed by law.’.
I would think this is a very complicated picture. I couldn’t guess the figure but many people who would have been murder victims 100 years ago, now survive due to medical advances. It might be more informative to divide the murders into different categories.
Fair point. But by the same token, what proportion of crime today [not including homicide] actually gets logged into a data registry due to better technology, storage etc. vs the past? We have significantly more advanced CCTV and forensics, so I’m wondering if people also report things more [since they have proof and police accountability systems greatly improved]. After all, a lot of folks will go on Twitter with their own footage if police do not take them gravely, something I suspect is taken more seriously today than in the past [due to things like social media outrage, populism and public shaming].
Age demographics, I would imagine are the biggest factor driving murder rates. Overall population figures only reveal so much. Murderers, like most criminals, cluster around the age range 18-35. In WW2, nearly a half million men died, mostly in this age range, so it seems to fit that murder rates would be low in the years following the war. A baby boom began immediately after WW2, peaking in the 60s, leading to an increase in murders from the 60s, peaking in the 1990-2000s.
Yeah, i cant believe they didnt correlate this to murder per capita or 100000. The kick up in murder followed the rising population. Exactly as you said, its not rocket science.
We’ve now added a chart for homicides per million population – if you compare the charts it appears to explain some of the rise after 1960 but the general trend is still there.
Except the rate of homicide increased from around 6 per million population in the early 1960s to 15.1 by the year ending March 2002.
Per capita increase in alcohol consumption kicks off around 1960, at the same time as the rise in homicides (https://publications.parliament.uk/pa/cm200910/cmselect/cmhealth/151/15106.htm). Since around 2003 it has begun to trend downwards (http://www.historyandpolicy.org/opinion-articles/articles/the-highs-and-lows-of-drinking-in-britain). This mirrors the trend in homicides since 1960. That there isn’t a decrease in homicides between 1900 and 1932, mirroring the fall in alcohol consumption may be largely due to the increase in population over this period. The homicide rate falls from around 1.05 per 100,000 in 1900, to 0.75 per 100,000 in 1932. By 1960 it has fallen further to 0.61 per 100,000. If we ignore the Shipman spike, it peaked at just over 1.5 per 100,000 in 2003, before falling back below 1 per 100,000 in 2012.
This is definitely something worth examining more closely, given the proportion of cases that involve alcohol in some way or the other. Not just because alcohol perhaps makes people more likely to resort to violence, but also because pubs and clubs that serve alcohol attract large numbers of people, increasing the chances of an incident kicking off that might end up in a homicide. There’s a study in Finland that suggests the drop in homicides there is mainly explained by a decrease in the number of young men killing each other in drunken fights (usually during nights out).
It’s also interesting that drug use appears to have declined during the same period late 90s to 2015 – suggesting that there is a wider explanation.
Per capita graphs would be a lot more informative. Including per-young-male-capita, as that is the group that most tends towards violence. Very hard to interpret absolute figures.
What I was wondering is – is the question that we have to explain “why was there an increase from the 1960s?”, or is it that we have to explain an unusual historically _low_ level in the 1950s? Is the post-50s rise actually just a case of regression to the mean and a return to the more usual level? Idly googling that is how I stumbled on this page.
Generally I have the sense that a lot of people mistakenly take their youth as the historic ‘norm’, when it may have been an era with its own very specific unique conditions.
A first glance at the population figures here: https://www.ons.gov.uk/aboutus/transparencyandgovernance/freedomofinformationfoi/populationbyagegenderandethnicity
doesn’t seem to suggest that per capita graphs would help that much though it would be interesting to look at given sufficient timé.
The only possible regression to the mean would be the 2003 to 2015 part, surely?
A simple way to check whether the death penalty is. Significant factor or not, is to introduce it in the UK ( following a referendum) and to observe what happens to the homicide rate for the ensuing years.
It states “But what caused that rapid growth in homicides from the 1960s onwards? Was it the state of the economy, the new ‘permissive society’, a breakdown of ‘family values’, or the effect of hard drugs like heroin and cocaine? Or a complicated combination of these factors (and others, such as the way the data is recorded)? And why has that trend reversed? Heavier sentences? Better policing?”
Without mentioning the abolition of the death penalty in 1965.
What are the statistics for murder only during the period 1960 to 2015?
What effect, if any, did the abolition of the death penalty have on these figures?
The trend in homicides appears to reflect the population growth, is this the case?
From a brief look at the population statistics there doesn’t appear to be any obvious correlation. (eg 1911 42m, 1961 52m, 2011 63m). The increase in homicides was mainly from 1960 until the late 1990s, and since then it has decreased.
The abolition of the death penalty wouldn’t explain the massive increase in all types of crime from 1960s onwards… or the decrease since the late 1990s. Obviously this doesn’t mean it wasn’t a factor, just that it’s hard to know whether it was a significant factor or not.