Last week I suggested at a meeting that the figures for unemployment in Britain had been manipulated. A member of the audience who said he was a civil servant was appalled and angry, suggesting that I was impugning the integrity of civil servants. I was somewhat taken aback by his outrage since I have become very accustomed, in the past ten years, to the thought that many government statistics are highly misleading. There are so many examples of it.
Hospital waiting lists are a prime example. Much unemployment is hidden under the category of incapacity benefit. Education is an outstanding example. My confidence in all government statistics has been completely undermined.
When I was a once-a-week leader writer for the Daily Telegraph (about five years ago) I would often start my research with some headline government statistics that appeared to support the official line and then find, on looking more deeply at the figures, that the headline figures were extremely misleading and, in some cases, that the real story was the very opposite of what the government was suggesting. Exam results are a well-known - or rather 'notorious' - example of this but there are many others that are less well-known and which succeed in fooling people (or at least the media).
As I explained to the irate civil servant, I am not suggesting that that figures are simply changed by the statisticians - that numbers are moved around Mugabe-style. No, I am suggesting something more subtle. I am also not suggesting that the statisticians are driving the misleading of the public. That is a political matter and therefore surely driven by the politicians and their 'special advisers' and public relations advisers - although let us not pretend that statisticians are all without political views and that all of them can put those views away when they select which, of the many ways of measuring things, they choose.
It is quite easy to manipulate statistics. One can choose the numbers that put the best gloss on things, ignoring ones which tell a different story. One can quote one study which ignoring others. One can redefine what counts and does not count as an instance, say, of an exam pass (just change the pass mark from 50% to 10% and you get a surge of apparent academic success.
One of the statistics I am currently suspicious of is the supposed fall in violent crime. I have not had time to mine the statistics. But I noticed this weekend a story in the Sunday Telegraph which encouraged my suspicion. A professor of 'advanced social sciences' surveyed frontline police officers, contacting them by email. No such study can be regarded as conclusive. However she apparently contacted 1200 of these officers which is quite a big sample.
She
...found that 80 per cent of borough police officers agreed or strongly agreed that knife crime was worse in their community than five years ago. Only eight per cent disagreed.Some 70 per cent judged that gun crime had worsened and nearly three quarters said they had seen a rise in gang crime.
Professor Qvortrup herself remarked that the result of her survey 'flies in the face of other research from the Home Office and the British Crime Survey, which says that gun crime is falling'.
I wonder. Is gun crime really falling? If so, why?
The full story is here.
Posted by James Bartholomew • Indexed in Behaviour & Crime
Comments (3)
TrackBack (12)
Comments
The problems with the crime statistics are very well laid out in the book Wasting Police Time - in terms of how the statistcis are actually collected and what is deemed a crime.
There is also another aspect of this, namely the refusal of the police to investigate crimes that are not on their targets - for instance, if you are a pedestrian that is knocked down by a cyclist on a pavement then there is absoutely no point in reporting this matter to the police because, though they may be sympathetic (and you may well be badly injured), they will not record it.
And you might argue that this is a minor incident and the police have better things to do with their time.
But actually it is this kind of thing that means people don't actually believe politicians when they claim that crime is falling, because their experience is entirely different.
Add to this that when serious crime rises - like the incidents of murder among teenagers in London - they are dismissed, or explained in some other way, that does not actually take into account the fact that the murder rate is rising - and effectively dismisses the issue.
In short the way in which the statistics are used is always to show that the qaulity of life is improving - which leaves people suspicious of the statistics.
Posted by: tfa at May 3, 2008 06:00 PM
James/tfa,
Your mention of statistics is something I should have taken up in my post.
Sophisticated statistical analyses of the crime data give scientific credence to the conclusions usually drawn – that crime is forever falling in our brave new world.
I have high confidence in statisticians and have never shared the common perception that statistics can be used to prove anything. However, scrutiny of the data, and data collection methods, almost always explains what is going on. The old computing adage rubbish in = rubbish out fits perfectly. For example, let’s consider murder. I see case after case in the media where one individual has taken the life of another, usually with intent, and yet they are not charged with murder, they are charged with a lesser offence. For other offences, the police and the criminal protection service often bring lesser charges, claiming that a conviction is thus more likely. Minor offences are increasingly ignored (does a caution get recorded for the statistics?).
And so the scam continues.
Posted by: John East at May 4, 2008 10:29 AM
Warning: file(http://63.247.138.2/~bartholo/randomquotes.dump) [function.file]: failed to open stream: No route to host in /home/bartholo/public_html/archives/2008/04/lie_damned_lies.php on line 327
Warning: implode() [function.implode]: Invalid arguments passed in /home/bartholo/public_html/archives/2008/04/lie_damned_lies.php on line 327
I have been puzzled by this as well. One explanation for apparently falling crime that I am sure is a big factor is the under reporting of crimes such as burglaries, or minor crimes of violence. Why bother reporting such offences when the police aren't interested and the courts take little action even if they convict.
However, the above argument cannot hold for gun and knife crimes because bullets and stab wounds must nearly always come to the attention of the police.
Could it be that the large increase in the prison population which has occurred over the last decade, in spite of all the efforts of the establishment, is the reason for a genuine fall in violent crime?
Those of us who have been campaigning for more use of prison rather than community punishments have always argued that locking more people up for longer times is the answer to rising crime. Perhaps this is what is happening, although I'm still far from convinced that this is the correct explanation. I don't think we have got tough enough yet to have produced any positive effect on crime levels.
No evidence for this either, but how about cheaper drugs and a more generous system of hand outs from Brown over the relatively prosperous last decade. This theory might be tested over the next decade with a deep recession on the horizon.
Posted by: John East at May 2, 2008 10:11 PM