Being rather late on the Horton Clockwise yesterday, Costcutter had no Guardians and no Telegraphs left, so I had to settle for a pinkun and the Independent. Rather to my surprise, the Independent contained quite a lot of stuff and has yet to be recycled in favour of what passes for news in the Saturday Telegraph.
So certain B-list celebrities on hard times have made a jolly good thing out of punitive damages for having their phones hacked. Much better paid and much less trouble than eating bugs in the jungle.
On the opposite page, there is a tale of the Home Office getting into a fix over putting the wrong dataset into the calculation of central grants to police forces. Which pressed all sorts of buttons with me.
I read that part of the calculation is a package called Acorn sold by a company called Caci and usually used by retailers rather than governments, but with the sensible common thread being the use of census data to classify areas. See reference 1. I remember Caci as being a very aggressive, IT & Oracle flavoured consultancy back in the eighties of the last century, a consultancy for which I was briefly tempted to work. So still up and running, although their offering seems to have moved on a bit and they are now very into marketing, surveys, branding and geography.
The basic idea of the calculation is common to governments across the world. You have functions organised by area, in this case the police. You get all sorts of data about those areas, you apply a funding formula to that data and so generate an amount of money for each area. So, in this case, areas with lots of drug takers might get lots of money. Areas with lots of unemployed young men might gets lots of money.
The problem seemed to be that the geeks plugged Acorn into the 2001 census - the census being all or at least a good part of the data in this case - rather than the 2011 census, which seemed to make a huge difference to the central funding for the Metropolitan Police. While the wonks did not think to wonder why there was such a huge difference - if indeed they noticed that there was one at all. Did they forget the first law of statistics that says that if a statistic looks interesting it is probably wrong?
But what sort of a calculation would change so much between 2001 and 2011? Do I want to use a funding formula for the police which is so unstable? The facts on the ground have not changed that much in that time. And what about reasonable expectations of the finance people in the police forces?
From where I associate to the problem of trying to be too clever, trying to be terribly fair and to take into account every relevant variable. Which can all too easily result in a formula which behaves in odd ways and delivers results which the areas, never mind the politicians, cannot understand. Rules for taxes and benefits often suffer from the same disease. Another manifestation, perhaps, of the best being the enemy of the good.
Is another strand in the problem the steady cutting back of headquarters staff in the Home Office, on the grounds that they are all unnecessary overheads, useless mouths to feed? Forgetting that one of their uses was to do a proper job on funding formulae. All part of the Tory wheeze of strip all the money out, then make huge 'how awful' noises when the inevitable mistakes get made, then privatise on grounds of incompetence. Let's let our friends make some money out of all the mess instead of all those dossy public servants.
I have even heard it suggested that large and once successful companies like Tesco's can fall into the same trap, cutting back their headquarter's staff to the point where they don't have enough accountants left to do a decent job on the accounts. From which flowed the scandal of the scoring of profits from pipelines of supply.
Back at the Independent, on the down side, there is an incomprehensible cartoon of a naked Cameron, very much in the style of, but not quite as offensive as the sort of thing which appears in the Guardian. Notwithstanding, perhaps I shall buy another Independent sometime soon.
Reference 1: http://acorn.caci.co.uk/.
No comments:
Post a Comment