Back in 2006, Tim Blair declared
I'd lean towards the official police figures myself (although jerked-around crime counting methods make comparisons problematic), mainly because they're, you know, official police figures. The British Crime Survey is just a survey.
Alas, he backed the wrong horse, as this story from the BBC proves:
Police miscount serious violence
A number of police forces in England and Wales have been undercounting some of the most serious violent crimes, the government has admitted.
It means figures for serious violent crimes rose by 22% compared to last year - rather than showing a fall as previous figures appeared to indicate. ...
However, the revised figures still show that overall crime and overall violent crime remain down on last year - a finding backed by the results of the British Crime Survey (BCS), also released on Thursday.
The BCS, which is a study of the experiences of victims, rather than police records, shows an overall decrease of 6% in the number of crimes compared to last year.
and:
Crime has been falling in almost every developed nation, regardless of the specific crime-busting approaches by different governments.
So while six out of 10 people think crime is rising nationally, the risk of being a victim is at its lowest level since the BCS began in 1981. The statisticians report goes on to note that those most likely to say crime has risen a lot are also most likely to read a tabloid. And that's a damning sideswipe at the media that ministers would never be prepared to make.
When he learned how unreliable the police figures were, tabloid journalist Tim Blair admitted he was wrong and posted a correction. Well, not exactly.
Wow, it's like reviewing a book by reading the title. You'd think that a defence of using police figures over BCS figures for trends would, you know, address at least one of the zillion reasons why the BCS is considered more indicative.
Perhaps the official police statistics accidently used hilarious crime to offset their serious crime numbers? An easy mistake to make.
This has been reported a lot here and I'm going to leap to the defence of the police here.
As I understand it, this issue relates to incidents where serious harm was threatened but no grievous bodliy harm was committed (eg. threatening someone with a broken bottle as opposed to gashing their face with it). Without any serious injury resulting from the incident it is usual for the offence to be prosecuted as some lesser crime, such as common assault.
Now, what the government have done is changed the goalposts and asked the police to record crimes that threatened GBH under the same serious heading as GBH, even though when those crimes proceed through the courts they would be regarded as lesser offences. This is all a part of trying to take unacceptable behaviour more seriously - extreme threatening behaviour arguably *should* be a more serious crime than common assault. But it's not a case of police historically fudging data or making a mistake in recording the crimes - they were recording the crimes with the seriousness that they would have been treated with by the courts.
The situation we now have is that the statistics will show the police are dealing with a higher number of serious crimes and fewer less serious crimes - but the numbers when followed through prosecution and conviction will remain unchanged. There is a disparity between the guidelines for recording crime statistics and the guidelines for prosecution.