Thursday, 4 June 2009

The rise in teenage murderers?

For some time now, I’ve been trying to verify a hunch that whereas murder was once an ‘adult’ activity, it has recently become more a preserve of the young —indeed, the very young.

About three years ago, I noticed the astonishing number of news stories online about murders in which the victim or the perpetrator was in their teens.

Often, if the victim was in their teens then so, it seemed, was the perpetrator. What was even more alarming was that in the case of teenage perpetrators of all kinds of murders there were often two or more individuals involved.

Trying to get specific numbers is difficult. Curiously, I have twice tried requesting information on the numbers of teenagers convicted for murder from our own Ministry of Justice, but without success (even though the website says this is possible under the Freedom of Information Act).

Today, however, I found that the Google news archive now shows a timeline of results for a search, and this is very interesting.

If you click here, you will see the result for the search term "guilty of murder" (this should open in a new window or tab, so you will need to come back to this page afterwards). At first glance this might seem to suggest the murder conviction rate has gone up since 1860, declined in the 1940s, and then has risen again since the 1960s. But of course what we also have here is the number of filed news accounts and a dramatically increasing population. On its own, the graph is interesting, but tells us nothing about rates of increase or decline.

However, if you click here, you will see the results for the same search term, only this time just for the period 1950-2009, and this is where it starts to get interesting.

Now click here, and see the results for the search term 'teenager + "guilty of murder"', again for 1950-2009. This time, we have factored out the increase in news media and population. The difference is, I'm sure you'll agree, striking. Click between the two pages or tabs to compare the two. Whilst the rate per head of population may not have changed, the rate of reports has increased dramatically. Indeed, between 1975 and 1995 it increased exponentially. If you scroll down the page, you will also see that the stories are, indeed, generally about teenagers who have carried out, and been convicted of, murders.

However, there may still be other factors involved, so as a kind of control try comparing 'wife+"guilty of murder"' and 'husband+"guilty of murder"' over the same period as before. This time it can be seen that both rates of reporting have increased - and the 'husband' rate slightly more than the 'wife' - but that the rates are not markedly different (and, thankfully, both seem recently to have fallen).

It is, I agree, not absolute proof. But it is surely an indication of a trend.

John Richardson
4 June 2009

When posting your comments please give a full name and location. Comments without this information may not be posted.


  1. From the 'We all going to hell in a handbasket' file?

    Too much time on your hands?

    Histograms with no scale, no numbers available, no controls for population number variation, not very scientific.

    Search 'Teenager' as a lone term and see a trend. Search 'Teenager succeeds' and see the same trend. You've discovered the term 'Teenager' is a more prevalent today than in 1950 not that they are committing anymore murders - on this evidence.

    Good luck with the FOI requests I hope they reveal more concrete evidence - and if it is contrary to your hypothesis that you'll publish it too??

    BTW if you type in 'pensioner' and 'guilty of murder', it may be me, but there looks to be the start of a 'trend' there too!!

  2. Donald, I'm well aware of the limitations of the 'method'. Like I said, I tried the reliable approach and it stalled. Interestingly, I actually got as far as having a phone call from someone who wanted to verify how I would like the data on teenage murder convictions, and yet nothing was ever sent! I'm not, however, completely naive about methodology. I have not 'proved' a trend, and am not claiming to have done so - I just think there is enough here, and in the news, to justify going on probing.

  3. PS to Donald, I'm not sure your "pensioner 'guilty of murder'" search quite looks as striking as the "teenager 'guilty of murder'" search for the same time period. Like I said in my post, you also have to check the actual news items listed to see whether they refer to incidents where a teenager is either murdered or is the perpetrator of a murder. Ie, it involves checking to see this is not just an increase in the incidence of the search term. It is also worth bearing in mind that there may actually be a real increase in the number of murdered pensioners and of pensioner murderers! Personally, though, I would still prefer hard data from the Department of Justice, so I will probably have to try again.

  4. What worries me is that you say you tried a more rigorous approach, the data isn't forthcoming, so you put your opinion out there instead. People won't remember that you didn't get the proper data to start with, just that you said there was a trend that teenagers were murdering more.

    I didn't say that the pensioner 'trend' was as striking - you have. I said that with no figures to the graphs, the 'trends' are meaningless, including a spurious look at homicidal pensioners or succeeding teenagers - in the eye of the beholder.

    You also don't say whether you'd comment if the data doesn't show what you want it to show - that is scary and damaging to your case of objectivity.

    As for taking news stories as a source of data here is someone else's view:
    From the ESRC Society Today
    Title: Unbalanced newspaper coverage is teaching the public the wrong lessons about homicide cases.
    "In the Times, for instance, between 1977 and 1999, just 13 cases contributed 2,860 stories out of a total of more than 15,000. Professor Soothill says: "As a responsible and credible source, newspapers need to highlight real dangers rather than manufacture inappropriate fears."
    Stories with a sexual motive were most likely to be reported (around 70% in each newspaper), followed by cases with a robbery or theft motive, and cases which appeared to be irrational acts....less than one in five cases where the homicide stemmed from rage of a quarrel was reported. Other important factors included the number of victims, their age and that of the suspect, the method of killing and whether a homicide involved a female victim."

    I too can't paste into the comments box - is it a Firefox thing??

  5. Donald, a couple of points.

    First, it is not quite true to say that there are, as you put it, “no figures to the graphs”. Nor does this mean that the trends are “meaningless” —an up is still an up and a down still a down. But in any case, the horizontal (time) axis is clear. Moreover, if you click on a section of the graph, you can then see the number of page results this represents.

    Regarding the 'teenager "guilty of murder"' search, for each half-decade from 1950 to 2009, the results are 11, 19, 12, 22, 34, 46, 100, 264, 539, 666, 835, 1460. You can also do the same on a ‘month by month’ basis. (By comparison, your ‘pensioner “guilty of murder” suggestion reveals a range from 5 to 85 over the same period.)

    You can then, if you wish, check individual news stories —which of course shows that some are about teenagers who murdered, some about teenagers who were murdered, and some which are just coincidences of the same words on the same pages.

    However, this in itself is not, I repeat, being claimed as absolute proof of my hunch that murders of and by teenagers are increasing. If you Google just ‘teenager’ there is a rising trend covering about the same period (this time, the vertical axis is about 636 to 462,000).

    Nevertheless, there are other sources. This ‘Smart Library’ article, for example, sums up a report titled ‘Trends in rates of violence in the U.S.A’, the synopsis of which says that, “Murder rates among teens skyrocketed during the 1980s to peak in the early 1990s. Since then the murder rate among teens in the U.S. has dropped.” This has ‘proper graphs’ to support it, and attempts an explanation of the trends.

    Or again, there is this American report: “Between 1970 and 1993, the homicide rate for teens ages 15 to 19 more than doubled, from 8.1 per 100,000 to 20.7 per 100,000.11 The rate declined dramatically during the 1990s, and has stayed between 9 and 10 deaths per 100,000 since 2000; in 2005, the homicide rate was 9.9 deaths per 100,000.”

    Or, from the UK, there is this: “Although both Metropolitan Police recorded crime and the British Crime Survey indicate that crime in London has decreased over the last five years, a tragic trend in teenage homicide has emerged. The total number of murders in London has been reducing since 2003/04 yet, despite this, the number of teenage homicides increased significantly in 2007, from 17 to 26. Furthermore, the problem has become worse during 2008, with 27 cases already between January and October.”

    One could go on, though the proverb about swallows and summers springs naturally to mind. The point is simply, like I said before, there seems to be something going on. My hunch is on a trend for the worse (though with some signs of recent improvement).

    The one thing I am convinced about is that its not just sensationalist reporting!

  6. PS to Donald, I like this article here which attributes the decline in violent crime in America to the Roe vs Wade legalization of abortion.

    According to the author, "The millions of women most likely to have an abortion in the wake of Roe v. Wade—poor, unmarried, and teenage mothers for whom illegal abortions had been too expensive or too hard to get ... were the very women whose children, if born, would have been much more likely than average to become criminals."

    Thus, "It wasn't gun control or a strong economy or new police strategies that finally blunted the American crime wave. It was, among other factors, the reality that the pool of potential criminals had dramatically shrunk."

    Darwin would have been delighted!

    The source is Freakonomics by Steven Levitt and Stephen J. Dubner.

  7. If we let it go that you start with a conclusion and are trying to fit the facts post-hoc.
    If we allow that conflation of 'teenager offenders' - the subject of your hunch, with 'teenage victims' - the subject of the Met Police 'evidence' you submit.
    If we allow your view of stats where 'up is up', in the face of the term 'teenager increasing five-fold more than the term 'teenager guilty of murder' over the same time span (your figures).
    If we ignore your assertion that the argument there is a sensaltionalising of stories in the press (evidenced by the ESRC) can be dismissed by your conviction.
    And finally politelly ignore your swipe at Darwin as I'm sure you know social Darwinism was never put forward by the man himself.

    I'm left wondering what the point of the social comment part of this blog is?
    Is it to test an argument - if you cannot change your position under weight of evidence it doesn't seem to be that?
    Is it to make the CoE readership feel better that their world view is correct?
    Is it to try and convert?
    Am I missing the implied political or religious aim?

    Or did you say it in another thread - 'it is my blog and it amused me'?

  8. Donald, I think the answer is, its my blog and it amuses you.