Comments on: A good day for civilisation

By: Phil

Phil — Sun, 24 Jul 2005 07:07:00 +0000

Not all the incidents reported to the police are recorded as crimes; not all the incidents recorded by the police as crimes are referred to the CPS; and not all the crimes referred to the CPS are proceeded with. Under Labour there's been strong political pressure to reduce the rate of 'attrition' at the first and second stages; the recording guidelines were explicitly changed a few years ago, from "incident is a crime if witness says so and there's some evidence" to "...if witness says so and there's no evidence against". While that made police figures leap up - making them effectively useless for comparisons over time - a more gradual trend upwards can reasonably be seen as reflecting continued informal pressure along the same lines. Add to that the effects of what happens when you introduce lots of police to an area with lots of pissed-up teenagers, and YES, OF COURSE POLICE VIOLENT CRIME FIGURES ARE UP - WHAT ELSE WERE YOU EXPECTING?

Sorry. All better now.

By: Peter Clay

Peter Clay — Thu, 21 Jul 2005 11:17:00 +0000

My opinion on this is that actual violence is down, but low-grade intimidation, abuse and street hostility which don’t register as crimes have increased, hence the rise in fear of crime. The Rowntree survey on ASB has some interesting commentry on this.

By: bystander

bystander — Thu, 21 Jul 2005 06:56:00 +0000

A great deal of the violence that gets to court is domestic, as there is a serious drive on to prosecute these cases (although conviction rates remain low). A generation ago the police would not get involved so long as the abuser stopped short of GBH, so the figures are skewed. Also, police now attend every domestic incident where violence is alleged, and I know of one woman who called police over fifty times in a year – that’s 50 crimes for the stats.

By: Matthew

Matthew — Thu, 21 Jul 2005 06:54:00 +0000

For the limits to this kind of stuff you should read John Birt’s "blue skies" document on crime for the PM (released under the FOI Act) (http://www.cabinetoffice.gov.uk/publicationscheme/documents/pdf/crime.pdf)

He attempts to calculate how many actual crimes take place, and comes up with 130m (10 times the BCS), of which 65m are cannabis offences. Not sure why he didn’t include speeding, as I would make that about 427bn offences.

By: Backword Dave

Backword Dave — Thu, 21 Jul 2005 06:27:00 +0000

"Do you happen to know if the police record separately (a) crimes reported to them and (b) crimes which result in charges being raised?"

Yes. They must. They issue crimes numbers for a). If charges are raised, these go to the CPS, don't they? So both are recorded. But crime numbers mean little. They're more like the numbered tickets you get for queuing in some supermarkets. If you report a spurious crime, for instance, you could still get a number.

Euan, uncertainly and greyness is the best we can do. See Theodore Dalrymple in the Times today.

By: Andrew Bartlett

Andrew Bartlett — Thu, 21 Jul 2005 06:19:00 +0000

Well, last night we has that guy from Westminster council on Newsnight banging on about how young people aren’t scared of police, how they know their legal rights (‘cos that is such a bad thing) and how the only way to counter this problem (after lamenting that the police in this country can’t whack people with trunchoens – for a guy who films city centres at night he must have a peculiar blindspot) is to draw up laws that allow the police to use a whole load of arbitrary powers.

By: Chris Lightfoot

Chris Lightfoot — Thu, 21 Jul 2005 06:17:00 +0000

But they aren’t "reporting the same thing", ostensibly or otherwise. The number of crimes reported to the Police is not the same as the total number of crimes that have occured, full stop.

By: dsquared

dsquared — Thu, 21 Jul 2005 06:16:00 +0000

They’re not ostensibly reporting the same thing. And the survey data is much less easy to manipulate than the reporting data.

By: Euan Gray

Euan Gray — Thu, 21 Jul 2005 06:11:00 +0000

Do you happen to know if the police record separately (a) crimes reported to them and (b) crimes which result in charges being raised? If not then this should be made clear. Ideally, we could see both, and there would need to be explanations of any big variances.

I'm suspicious of this survey approach to establishing empirical data. It plays into the hands of those who would say that we are going to Hell in a handbasket and the only solution is to arm the populace, precisely because it is way too easy to cast doubt on its accuracy. It's also susceptible to political manipulation, although I'm not suggesting this is necessarily happening. I don't think it's meaningful to call it anything more than a general impression of the level of crime, not the actual level of crime. It may be accurate, it may not.

Whilst popular opinion about crime levels is important to know, it is also important to know the truth about the actual numbers of crimes happening.

I don't see how anyone can really be confident where there is an 18% point difference between two figures ostensibly reporting the same thing.

By: dsquared

dsquared — Thu, 21 Jul 2005 06:05:00 +0000

The BCS is carried out by Mori, using their standard sample selection methods. I think that there are reporting method issues; the standards changed during 2003 so 2004 would have a full-year effect.

I disagree with John about pretend coppers, btw; Camden was, as with so many other things, an innovator in this one and they don't actually do a bad job. Lots of them are nightclub bouncers in the evening so they know where the bad lads are.

In general I would advise people that living your life not in fear is the best way forward. Only the other night the missus mentioned how polite our local Somali crack gang are, and how they always move off the pavement and say good morning when she needs to get the baby's pram past.