Who killed murder?

A detective writer lines up the suspects on the mysterious worldwide decline of murder and robbery

19 March 2016

9:00 AM

19 March 2016

9:00 AM

Pity the poor crime writers. Our earnings, like those of all authors, are diminishing for reasons far beyond our control. Our fictional criminals and detectives are being outsmarted by genetic fingerprinting, omnipresent security cameras and telltale mobile phones. Who needs Sherlock Holmes to solve a tricky crime when you have computers, with their unsporting ability to transmit and analyse enormous quantities of data and identify culprits? But the bigger problem for us novelists (if not for everyone else) is that murder itself is dying.

The official homicide rate peaked in 2002, thanks to Dr Harold Shipman, and has since fallen by half — from 944 then to 517 last year. Adjusting for population, murder is now at the same level it was in the last years of Queen Victoria — and, in spite of what Arthur Conan Doyle led readers to believe, the streets were pretty safe then. The murder rate started to rise in the 1960s and soared in the 1990s, which caused widespread panic. Family breakdown, collapsing morals and a feral underclass were all blamed for an apparently inexorable increase in violent crime.

But then that rise stopped, and the murder rate began to nosedive. Why? Or, as we crime writers say, whodunnit?

Investigating the murder of murder is difficult, especially when statistics derive from a variety of sources. This is not just a British phenomenon. Homicide rates throughout the industrialised world have declined during the past century. The fall in Britain, the USA and Canada is particularly marked — these are all countries where the rate increased in the 1960s and 1970s. And it isn’t just murder. Over the past 20 years robbery has declined almost as much as homicide, and vehicle theft more so.

Prosperity is often fingered as the prime suspect, the idea being that the economic boom of the 1990s suppressed many of the conditions in which crime flourishes. It’s a nice theory, but unfortunately it doesn’t hold up to scrutiny. You can search in vain for any correlation between national wealth and crime figures. During the Great Depression, when unemployment rose to 25 per cent, the crime rate in many cities went down. When Britain became far richer in the 1960s, crime started to spike. But since the beginning of the economic crash, in 2008, murder rates have continued their downward trend.

Evidence from the rest of the world further undermines the idea that wealth is killing homicide. In its 2011 Global Study on Homicide, the United Nations Office on Drugs and Crime confirmed that the murder rate had fallen worldwide during the recession.

Which leads us to another possibility: perhaps the murder rate has fallen because we have put most potential murderers behind bars for other crimes? Britain’s prison population has doubled over the past 25 years, and almost doubled in Australia and the United States. Alas, this theory is fatally undermined by conflicting statistics from other developed countries, such as Canada and the Netherlands, where they’ve released plenty of prisoners without any corresponding surge in homicides. Even in New York, once a hotbed of murder, imprisonment has fallen by 26 per cent over the last decade — and crime has fallen by 28 per cent.

So we must search for another lead. Perhaps the police are getting better at preventing crimes? If this is true, the police must have brought about the reduction in violent crime with fewer officers: last year, Britain had 154,000 boys and girls in blue, some 10 per cent less than when David Cameron came to power. For the Police Federation, who were arguing against the cuts, it’s all rather embarrassing: their budget was slashed and surveyed crime is now at its lowest since records began 25 years ago.

The increase in ‘data-driven’, ‘CompStat’ or ‘hotspot’ policing has no doubt helped prevent a lot of crime. This involves the flexible use of police resources, allowing forces to respond rapidly to potential trouble spots by flooding them with officers. Another tactic is the so-called ‘broken windows’ approach, whereby police crack down rigorously on low-level offences in the hope of preventing more serious ones in the long run. A third method involves building trust through community–oriented policing.

But is any of that really the difference between a rising crime rate and a falling one? The evidence from America is mixed. One study looked at three cities that each pursued a very different approach to policing. In New York, the police force quadrupled in size and there was what is politely called an ‘oppositional’ relationship between neighbourhoods and officers. In San Diego, the number of police on the beat remained much the same as before. In Washington DC, the police chief pursued an ‘empathy’ policy, reducing the level of confrontation between police and public. Three cities, three policies, and in each case the result was a swift drop in crime.

Some experts argue that the decline is due to wider demographic changes. Western societies have more old people now, and the elderly are much less likely than younger people to commit violent crime — or to be victims of it. The original baby boomers are collecting their pensions, and the second generation of baby boomers are now in their forties. But in London this answer doesn’t stand up to close scrutiny either. The number of 18-to-24-year-old men — the group most likely to turn to crime — has increased, while London’s crime rate, apart from a small spike in murder last year, has continued to drop.

So perhaps it’s just that the usual suspects — young people — are better behaved than their parents were? They are more likely to be university-educated and, across the EU as a whole, a quarter of them aged 27 to 34 are still living with their parents. They are less likely to drink or consume illegal drugs than their parents at the same age. Sober young people tend not to kill.

Some think that the real clue lies in the womb. In 2001, an American economist named Steven D. Levitt argued that legalising abortion may have had the unexpected side-effect of lowering the crime rate. This controversial theory is based on two premises: that unwanted children are more likely to turn to crime; and that legalising abortion means fewer unwanted births.

A number of American and Canadian studies have established the possibility of a link between legalised abortion and diminishing rates of homicide. If they’re right, Levitt claims, this would account for 25 to 30 per cent of ‘the observed crime decline in the 1990s’. Correlation, however, does not amount to causation and the theory does not appear to work outside North America. The tapering off of criminal activity has also continued long after abortion rates stabilised.

And so we turn to identifying the murder weapon. Countries with strict gun laws have few homicides. Among OECD countries, America’s murder rate is second only to Mexico’s, though the number of violent assaults is comparable to other western countries. Guns are used in two thirds of all murders. Since Obama has been in the White House, however, gun sales have actually risen. In 2013 alone, Americans bought about 16 million new firearms. Yet the rate at which they gunned down their fellow Americans has halved over the past two decades.

Of course, it is worth looking at the lifestyle of potential perpetrators. The street price of heroin has plunged in recent years (it’s now about £10 a hit), which has led to suggestions that a fall in violent crime is at least partly due to the fact it is now easier for addicts to fund their habit without crime. Similarly, the crack epidemic’s decline has been cited as a contributing factor.

No theory, however unlikely-sounding, should be dismissed without thorough investigation. And so it might be that the phased removal of lead from petrol and paint caused a drop in violent crime. Lead pollution has a well-documented effect on children’s brains, leading to aggressive behaviour and cognitive delays. Some researchers have matched the rise and fall of such pollution to the rise and fall of murder, though oddly enough they have found no link between it and property crime. Nevertheless Jessica Reyes, an American economist, credits cleaner air with 56 per cent of the reduction in violent crime. Many other experts find the theory suspiciously simplistic and less than plausible, so in the absence of supporting evidence it’s not convincing. The increased use of mood-enhancing drugs such as Prozac and Ritalin has also been suggested as a factor.

If this were a murder mystery, it would be the unsatisfactory sort that has too many far-fetched suspects. The solution in one of Agatha Christie’s better-known stories turned out to be that all the suspects were guilty of the murder. Perhaps that’s happened here. But we really don’t know. The creator of Hercule Poirot would have arranged things much better.

Got something to add? Join the discussion and comment below.

Andrew Taylor’s latest novel is The Ashes of London. Andrew Taylor is a winner of the Cartier Diamond Dagger for lifelong excellence in crime fiction, and a regular crime reviewer for The Spectator.

You might disagree with half of it, but you’ll enjoy reading all of it. Try your first 10 weeks for just $10

Show comments
  • outlawState

    The State killed it. As we move ever onwards to the Orwellian big brother society, the crooks are more likely to be wearing police uniforms, in the government, and be “enforcers” of all kinds. The oulet and opportunities for State sponsored thuggery are ever on the increase.

    • OmnipotentWizard

      Quite right – it is important that murderers have more freedom to commit their crimes.

    • Father Todd Unctious

      Well the state authorised divorce, abortion, contraception and education……so yes the state killed it.

  • Jonathan Tedd

    Human nature doesn’t change – murder is primarily a domestic affair motivated by lust, greed, anger…I highly recommend this book:

    A Companion To Murder: An A-Z of Notorious Killers and Sensational Trials.

  • Polly Radical

    Surely the availability of the Internet from the early 1990s is the correlation.

    When Internet-based crime is so easy, profitable and low risk, why should criminals leave the house at all?

    • Father Todd Unctious

      You attribute a worldwide decline in violence and murder, that has gone on since at least the 18th century,to the appearance of the internet in the 1990s.

      • Polly Radical

        Are you suggesting that people in ‘developing’ countries are inclined to murder???

        • Father Todd Unctious

          No. I imply that educated folk are less murderous.

  • terence patrick hewett

    George Orwell did it in 1946 with his essay: The Decline of the English Murder.

  • Roger Hudson

    What bull****, Take your 517 murders and strip out the ‘domestics’, the drunken rages etc. and you have hardly any left. Those murders that were ‘capital’ after 1957 are very few.
    The law should really change and have an American style ‘degree ‘ system. For me first degree would still be the crimes from the 1957, capital murder list, with the same penalty.

    • enoch arden

      The death penalty in the US became an idiotic farce. For about 14000 homicides a year, only 38 were executed in 1914. And these executions were performed after years of absolutely schizophrenic legal procedures which cost millions in each case. At the same time, police kill several people every day, without being responsible. Including 12-year old children.

      • Jackthesmilingblack

        So the biggest killers in the U.S. Are the police. No surprise there.k

  • enoch arden

    There were 325 murders in 1965, right before the death penalty was abolished. The humanists who pushed for the abolition are responsible for the thousands of extra murders committed within that interval of time.

    The UK should follow the example of Singapore.

    • Father Todd Unctious

      In 1965 we had 40 million adults. Now we have 51 million. Most murder is committed by adults. We have 28 % more adults so expect 28% more murders.
      I put the decline down to the coming of Abortion in 1967 and the widespread use of the Pill after 1974. About 25 years later one would expect to see fewer crimes all round as many of the previously potential criminals simply would not have been born.
      A lot of murder was domestic , so increased divorce rates would help massively too.

    • CocklecarrotJ

      The evidence simply does not support that argument.

      There is plenty of research available on the effectiveness of the death penalty as a deterrent (and I’ve read a fair amount of it, because I did a degree dissertation on it).

      The conclusions are mixed: some data shows a positive correlation between capital punishment and reduced homicide, some no correlation, and some a negative correlation.

      Regarding the UK, in 1965 the murder rate was 6.5 million per year. It is correct that the rate is higher than that now, but you cannot therefore imply causation. For example, your statement could explain fluctuations in the rate over time when the existence/lack of capital punishment has remained the same. As the article says, these are material changes and cannot be explained on your argument.

      • Enoch Powell

        Remember that justice has a number of purposes. Working as a deterrent to criminals, providing closure to victims, punishing a crime, rehabilitation of minor criminals. The current justice system achieves pretty much none of these things with all criminals except lifers serving only half a given sentence. Even worse, it now allows for Islamic radicalisation in prisons as they are full of terrorists we can’t deport. What’s worse than Levi Bellfield, evil multiple murderer? Levi Bellfield, Jihadi!

        When the death penalty was abolished it was only achieved because the life term penalty was offered as a replacement to keep the worst of the worst off the streets. That has not been honoured.

        I quite agree that the death penalty is not a deterrent. It would however help victims achieve closure (and if a victim wished to forgive, the Judge could offer the victim the ability to commute the death penalty to a life sentence, meaning the criminal would really see the consequences of their actions and those who might be saved might actually feel shame) and it would punish the crime. It would also free up space in prisons allowing lesser criminals to serve their full sentence and prisons to enact rehabilitation properly instead of spending all their time attempting to manage overcrowding with all the attendant risks that brings.

        • enoch arden

          I would like to correct you on two points.

          The purpose of punishment is not rehabilitation. It is too late to reeducate a murderer. By hanging the latter we educate the rest of the population by the fear induced by this punishment.

          The second point was expressed 2400 years ago by Shang Yang. Small crimes should be punished severely, then big crimes will not appear. If big crimes appear in multitude, punishment becomes useless. The latter situation must be treated by imposing emergency situation, suspending constitution and using army for shooting criminals on site. That is, by terror.

      • enoch arden

        The evidence perfectly confirms that argument. Besides the UK statistics which you qualify as “fluctuation” (40-years long one) , I presented even more convincing evidence from Ireland.

        But the main empirical proof is Singapore which used to hold 2 world records: the highest rate of hanging and the lowest crime rate. It is logically clear that if the probability to be hanged for a murder exceeds 50%, the homicide rate will drop dramatically: killing will be an act of suicide. Criminals are not religious martyrs: they kill in order to live, not to die.

        • simon

          That is only potential empirical proof that that policy works in Singapore’s own social/economic/cultural/demographic environment. That does not necessarily mean it would work, or should be implemented in another country 100’s of years later.

          The list of influencing variable from one country to the next a swell as time period are huge really.

          • enoch arden

            Could you present an opposite example when the crime rate (murder rate) decreased after abolishing the death penalty?

    • Father Todd Unctious

      So are the IRA, the UDA and Harold Shipman.

  • barbara thomas3

    Julian . I just agree… Randy `s stori is unimaginable, last friday I got a great new Alfa Romeo when I got my check for $4136 this last five weeks and in excess of $10,000 last-month . no-doubt about it, this really is the easiest-work I have ever done . I started this 8-months ago and immediately began to earn over $77, per/hr . check out this site,……….….http//EmployeeAppreciationDay.org/GetPaid/98$hourly


  • al_frick

    Most likely the broken windows policy. In 1970’s New York, the streets were grimy, windows were broken, debris everywhere. Criminals who were wont to commit crimes saw this general chaos and figured there was no authority present. Even non-criminals littered and jumped turnstiles because the general aura was one of disorder.

    With prosperity came cleaner streets, shops that weren’t shuttered, street lights that worked. It is well-known that a simple street light or false camera can dissuade a criminal from a random criminal act. This is why NY puts in temporary police towers in the worst neighborhoods, most of them completely unmanned. Their deterrent effect is confirmed.

    Starbucks, higher end retail all contributed to a feeling of order.

    So I would say it’s a combination of broken windows policing, prosperity, and generally cleaner streets and order.

    • al_frick

      I know this first-hand. Take a micro-sample: Union Square in NYC. Used to be full of drug dealers, criminals, the worst element imaginable. When the 90’s economic boom took place, all the cheap discount stores were replaced with high-end retail Starbucks, Barnes & Noble, Whole Foods. NYU took over the western side of the square. High end restaurants sprung up. Eventually the park was cleaned up as well. Criminals were priced out.

      It’s more difficult psychologically for a criminal to mug someone in broad daylight on a clean, prosperous street than at night in a run-down, boarded up part of town. For the same reason as it’s more difficult to kill someone while looking them straight in the eye than shooting them in the back.

      • Father Todd Unctious

        So not the fact that by 1995 hundreds of thousands of potential criminals had not been born since Roe v Wade.

    • Father Todd Unctious

      You attribute a worldwide decline in murder to tidier streets in New York.

  • LoveMeIamALiberal

    Demographics is undoubtedly a major factor in all countries where the post WWII baby boom was a factor; the increase in crime from the late ’50s to ’80s is consistent with an increase in the number of young males. The chances of being caught have increased since the introduction of DNA testing in the mid 80s, which has deterred some calculating would be criminals. And improvements in medical treatment mean some attacks that would have resulted in death now do not, so they are recorded as assaults rather then murder.

    What is harder to understand is a change in young people’s behaviour; the generation since 2000 are less likely to drilnk, some and take drugs than their predecessors. Perhaps it’s the 60s progressive generation who are the historical blip?

    • Father Todd Unctious

      Not so for the UK. We had a postwar baby blip from 1944 to 48, then a significant lull in biths until 1957. Not sure where the extra young men you assume roved the streets in the late 50s came from. They would have been born in the 30s.
      Our boom was 1958 to 1972. Those young men came of age as criminals between 1980 and 1995. Most important is those who were never born after 1972.

      • LoveMeIamALiberal

        15-25 is the peak age for male criminality so births from about 1944 would start to have an impact on crime levels. Births during the lull of 50-57 were still greater than in the 30s. With births declining after 72 one would expect crime rates to fall late 80/early 90s, and indeed they did. It’s a very rough correlation to be sure.

        • Father Todd Unctious

          Peak age for crime, but not murder. The average UK murderer is 34. So for murders in the late 50s we look to those born around 1922. There was a huge baby boom after the Spanish flu epidemic in 1921/22.

          • LoveMeIamALiberal

            Murder numbers in the UK did not start a consistent upwards trend till 1965, which would be consistent with a baby boom c1930. The upward trend coincides with the abolition of capital punishment.

          • Father Todd Unctious

            Not strictly true. They hit 400 in 1952 and fell to 260 in 1958 rising thereafter. They also fell again between 1974 and 1983.
            The peak years 1993 to 2007 exactly mirror the babyboom chort hitting age 35.

          • LoveMeIamALiberal

            Misleading. Murders in 1964 were no greater than in 1930. It was only from 1965 that they rose above 300 a year and kept on a rising trend for the next 40 years.

          • Father Todd Unctious

            Not misleading. Fact. You are choosing to be selective with statistics. For England and Wales(N Ireland distorts figures after 1969) there were only 2 years between 1942 and 1957 with under 300 murders. Four years saw at least 370.
            The figures are easily obtainable from the ONS. So 1965 was far from the first year with more than 300.

          • LoveMeIamALiberal

            I think we are looking at the same numbers. Murders in 1964 at about the same level as 1930 (300 a year), variations in the interim, up to 500, but it is only since 1965 that there are been an upward trend taking the number of murders to previously unseen levels and remaining way above the long term pre 1960s level.

  • enoch arden

    Here is the Irish statistics of homicide rate. The death penalty was abolished in 1964. Look at the effect:


  • Owen_Morgan

    Daft article. I don’t ordinarily look to the Guardian for support, but…

    “The number of homicides in England and Wales rose by 71 to 574 in the 12 months to September 2015 – an increase of 14% fuelled by rises in knife and gun crime, official statistics show.”

    Do twelve-month statistics matter only when they date from January to December (and are the 2015 figures genuinely in yet?)

    Apart from anything else, it’s blindingly obvious that the police, the judiciary, the Home Office and the CPS have been shamelessly fiddling the figures for years, by disguising clear murder cases as manslaughter.

    • Father Todd Unctious

      Only 45% of murders end with a murder conviction. Of those, 15% serve 9 years or less. Shocking.

  • Michael H Kenyon

    Most murders are, historically, violent conflicts between people who have little, fighting to improve their apparent status or resources. If basic needs are met, or resources are acquired in different ways (e.g., courtship via “Tindr”), fights outside the “Dog and Duck” or the “Taj Mahal” regarding a girl or a spilled pint reduce. What’s left are murders for less prosaic reasons, which are less common generally.

  • Oli Norwell

    One other factor that I’m very surprised nobody has mentioned. The ready availability of entertainment and distraction in the form of smartphones, tablets and computer games.

    In the late 80 you had bored kids wandering around neighborhoods looking for something to do… Those same kids 30 years on are far more likely to be indoors messing around on their phones. Young people are getting a ‘fix’ of excitement/stimulation from technology and are finding they no longer need to mess around in the local area looking for trouble.

    • Oli Norwell

      … This all meaning that people with the need to satisfy violent tendencies are doing that in their bedrooms on the latest game rather than getting into fights.

      This all leads to less community violence and ultimately less murder and serious crime.

      • Father Todd Unctious

        But then,that is not the case.

    • Father Todd Unctious

      But kids do not commit murder. The average murderer is aged 35. You suggest that in the past bored teenagers set about killing people for want of entertainment. How ridiculous.

  • Chris Bartelt

    Unleaded fuel?

    • Jackthesmilingblack


  • Jackthesmilingblack

    Woman Murdered: Honest Jack
    The husband, 2/1
    The boy friend, 3/1
    The landlord, 7/2

  • Jackthesmilingblack

    So shocked to find a story where I’m not blocked, I can’ remember what I was going to say.
    Oh, yes. That bombing in Belgium, got false flag written all over it. Was these a training exercise running in parallel?

  • David Booth.

    The Authorities are massaging the statistics for murder by reclassifying them. Here in Scotland we’re told the murder rate is around 27 per year and yet not a week goes by its reported that someone meets a sad demise at the hands of his fellow man.

  • simon

    Of the list of things raised in this article that could impact on murder rates the likelihood is that any and all of them are to blame for rises and falls in rates – not just one thing, which seems to be what many seem to think is the case. Those influences will have had differing impacts at different time in different combinations over different periods across differing demographic pools in time in different countries (and even within countries with larger populations and/or differing regions with their own laws – The US for instance).

    Anybody that has ever been trained in the social sciences will understand this. It is currently a bug bear of mine that -for instance – many in the US pro/anti gun control lobby appear to think that the _only_ thing that impacts ‘violent crime’ stats and gun crime is the gun control policy of the country.

    Anybody that takes a country like Honduras ‘compares’ it’s gun murder rate to a country like Switzerland and then suggests that the only reason for any differing crime rates is that one has more (legal) gun control than the other really needs to start researching in a bit more depth than just grabbing “stats” from countries they otherwise no next to nothing about.

    Crime is a social issue it therefore requires far more qualitative research to fully understand it – not just quantitative stat comparing – often between countries with little influencing factors in common.