Crime stats - go figure!

[greentown]

Saw this story in the Independent -

www.independent.co.uk/news/uk/crime/what-riots-crime-stats-ignore-summer-disorder-2366152.html

Can anyone tell me what the theory might be behind this practice?

I think the article explains it. They give the example that the way crime data is captured for the map means that 10 people attacking one shop is counted as one incident rather than 10 crimes. I suppose that's a product of trying to distill complex information into a simplistic format.

Hm... I suppose if a business is robbed, it's one crime whether it was done by two people or a gang of 50.
You'd have to put these figures together with the arrest figures and charge/conviction figures over the summer to get a true picture.

Also, outside the hotspots of Hackney, Croydon, Clapham Jn and a few others, there was a lot of stuff called 'rioting' that wasn't - down my high street the media tried saying 'rioting has spread to X', but then they interviewed a bunch of business owners who all said 'no it isn't, this is just a bit of typical weekend vandalism that's just getting bigged up'. It's quite likely there wasn't that much 'extra' crime compared to a typical Saturday night of people out on the piss (especially as many of those would have stayed home to avoid it).

So if nine people were involved - let's specualte that nine individuals each stole a pair of trainers for example - then nine individual crimes would be recorded.
However, as soon as ten people steal ten pairs of trainers, then only one crime is recorded - is that the gist of it?

Can quite see the logic of that - perhaps I'm just being picky though.

The way the article explains, if nine people attack nine different shops, that's nine 'incidents' for the crime map. If nine people attack one shop that's one incident recorded on the crime map. But that's only the crime map - a simplified snapshot. As notcitrus explained, there are lots of other statistics published, so the crimes are far from 'hidden'

Crime stats are generally produced to suit the government. The method of collecting data I think varies. As the article highlights, 10 people robbing a store should be ten offences, not one. Any stats should have 10 to 20% added and then there will be a better reflection of the reality.