Monday, October 20, 2008
Why Nobody Trusts The Crime Stats
According to the government, crime is falling. According to the evidence of our own eyes, it isn't. One of us must be wrong.
The government's line - enthusiastically supported by the Prog Con - is that people who deny the fall in crime are rabid right wingers wilfully ignoring the evidence. Mark Easton, David Aaronwitz and Danny Fink all assured us of that after the latest official stats were released during the summer (see previous blogs here, here, here, and here).
As we've blogged many times (eg here), we do accept that some crime has almost certainly come down. Property crime - burglary and car theft - has fallen as anti-theft technology has improved and the market price of nicked DVD players has slumped. But we wonder how that will fare in the recession, and anyway that's not the crime most of us really worry about (and btw its fall is no thanks to the government).
When it comes to violent crime - crimes against the person - the crimes we really do worry about - we simply don't accept the government's account. We don't accept the crime stats.
We think the police recorded stats are now routinely gamed and distorted just like any other tractor production numbers. And we think the British Crime Survey is nothing more than an opinion poll commissioned by people (ie the Home Office) who have A Very Big And Very Obvious Axe (see blogs here and here).
So how big are the distortions?
Yesterday the Sunday Telegraph published some numbers on gun crime it has assembled via FOI requests to individual police forces. They say:
"The number of firearms incidents dealt with by officers annually is 60 per cent higher than figures stated by the Home Office.
Last year 5,600 firearms offences were excluded from the official figures. It means that, whereas the Home Office said there were only 9,800 offences in 2007/8, the real total was around 15,400. The latest quarterly figures, due to be released on Thursday, will again exclude a significant number of incidents."
They also remind us of their similar findings on knife crime:
"Knife crime figures were at least two-thirds higher than official figures. Police statistics showed forces in England and Wales are on course to record 38,000 serious knife crimes this year, or more than 100 a day, compared with last year's official total of 22,151 offences, a figure announced by the Home Office in July in its first annual count of knife crimes."
So the government's official crime figures understate police recorded gun crime by 60%, and knife crime by two-thirds.
Remember that next time the Prog Con asserts the crime stats are correct. They're not.
PS What would we do? First, as we've said before, the Home Office should not be marking its own scorecard - all national crime stats should be transferred to the ONS soonest. Second, our elected local sheriffs would reconnect policing with local communities, where disembodied official stats would cut much less ice than the actual experience of people on the streets.