Custodial Sentences

James Slack (of worst in the EU fame) has been at it again…. Although  to be fair, he so nearly made it through his piece on the almighty mess being made of the DNA database without looking too daft… if it wasn’t for one pesky pop-out bubble thing (what are they called?)

Almost 13,000 people a year are imprisoned on suspicion of crimes they didn’t commit, according to research..

Now of course, tradition would have it that the prisons are completely full of innocent people (reminds me of the anecdote of a relatively well known criminologist asking inmates if they’d have committed the crime if they thought they’d get caught for it); but despite the general ‘I dun nuffink’ approach, surely this is a bit worrying….

… according to Sheila Bird, a statistician with the Medical Research Council, in a paper for the straight statistics web-site…

It’s also worth looking at the source material , here it is… a nice piece from Prof Shiela Bird, of Cambridge, no less… Now, far be it from me to grumble about a programme director of an incredibly impressive research institute (again)… but I find the following:

How can courts’ performance be justly improved?

Rather sticks in the throat, – as two seconds on any well known search engine will tell you, the reasons for remanding people to custody is as follows

• The individual has been accused of committing a particularly serious offence.
• The individual has previous convictions for similar offences
• There is reason to believe that the individual may disappear before their trial.
• There is reason to believe that the individual may interfere with witnesses.
• It is likely that the individual will commit further offences before their trial.
• If an individual did not obey the terms of a previous community sentence, then custodial remand rather than bail maybe imposed.

(Prisoners Families Helpline – Google Hit #1 on “remanded in custody”… so we’re not talking deep criminological theory or Athens locked articles here)

Hadn’t we better have the discussion about whether this is actually a bad thing or not, before setting ourselves some (more) performance targets?

(As an aside, it amuses me greatly that the Mail is tacitly aligning itself with research which suggests we should have more suspected criminals on the streets)

I think the most depressing thing about this is that the overall aim of the Straight Statistics bunch seem fairly laudable…

• Draw attention to the inaccurate reporting of statistics in the media, and encourage better standards by training and example.
• Ensure that Governments follow the Code of Practice on the use of official statistics and the guidance issued by Sir Gus O’Donnell in February 2009.
• Monitor the use by local government, advertisers and industry of claims based on statistical or numerical information.
• Examine the statistical basis of claims made in scientific and medical journals, in the interests of greater clarity and sounder reasoning.

But the overly-politicised writings of their researcher coupled by a tabloid hack end up making the whole lot look pretty shambolic really.

The Place Survey – Some Observations

Back in July CLG finally got around to booting out the Place Survey (The new bi-annual ‘tell us about your local council and local area’ exercise) results to near-universal disinterest from the mass media. And why this lack of interest? – possibly something to do with the fact that the results have been released over 6 months after the fieldwork closed – And the reason for this delay?

Rumour has it that the ONS threw their rattle out of the pram 48-hours or so prior to publication in February because they didn’t really have any idea whether it was methodlogically sound or not…

On the advice of the review, the provisional data were revised. (p14)

… it wasn’t then…  Primarily this related to the issue of weights, the statistical application of underlying population data (age, sex ethnicity, household composition etc.) to survey data in order to remove error brought in by certain proportions of the population being more likely to respond than another… More on the review can be found on page 14 of the above document.

Anyway looking at the now deeply fiddled-with (but with a happy National Statistics stamp… so that’s fine, then) results it appears that the big headline (well, I say big), is that people are less satisfied with their councils than they used to be… Now it may be deeply cynical to observe that the questionnaire itself places the question

Thinking generally how satisfied are you with the way <your council> runs things?

Straight after

To what extent do you agree or disagree that <your council> provides good value for money

( core questionnaire – CLG)

I’m not entirely sure it takes a background in questionnaire design and response bias to think that perhaps this question may impact on the following question, particularly in a recession (Although, it was still the Credit Crunch back in October, wasn’t it?). I’m sure that, like last time, there’ll be a big project examining why it fell so far. (Anyone got any idea how much was spent on The Reputation Project* work? Audit Commission have been very quiet about it… funny that… still, it kept the sainted Ben Page in new suits, so who are we to complain.) .

And of course,  no doubt many local authorities, as we did, conducted their own analysis and spent quite a bit of time on it; just to discover that the official figures had been fiddled with… Waste of Time… Value for money? – not ‘arf.

*Trans: Shit, we don’t know what’s going on  – better throw some money at MORI and get them to spend a couple of hours on excel.

The League of Shame…

Never mind reported crime, we can make up statistics without waiting for the ONS to validate the Home Office’s fiddlings…

It’s been well documented what a sorry piece of journalism was inflicted on us by the Mail, Telegraph et. al. in their reporting of some international crime hysteria. For those of you fortunate enough to have missed it, the basics of the whole thing are best summed up by the following quote from the Telegraph:

The UK had a greater number of murders in 2007 than any other EU country – 927 – and at a relative rate higher than most western European neighbours, including France, Germany, Italy and Spain.

It also recorded the fifth highest robbery rate in the EU, and the highest absolute number of burglaries, with double the number of offences recorded in Germany and France.

Overall, 5.4 million crimes were recorded in the UK in 2007 – more than 10 a minute – second only to Sweden.

Even before entering in to the vaguaries of pre and post NCRS crime reporting rates and the way it throws comparative reporting into question (because, well, it’s really rather boring)  it’s worth having a look at Eurostat’s own 2009 study on comparable Crime and Criminal Justice rates across the EU.

It is possible to compare trends in total crime
over the period 1998-2007 for only about half
of the EU Member States.

less than 5 minutes checking the source of the article raises a report which proves there to be fundamental flaws in the article (page 11 of the report goes in to far more detail about exactly why the data shouldn’t be compared in such a way).

Now rather than lie, we get some wonderfully vague statements of method….

Researchers admit that comparisons of crime data between countries must be viewed with caution because of differing criminal justice systems and how crimes are reported and measured.

(from Richard Edwards at the Telegraph, see above for ref.)

or

But criminologists say crime figures can be affected by many factors, including different criminal justice systems and differences in how crime is reported and measured.

(from the aptly surnamed James Slack at the Mail)

But somehow I doubt that sort of statement gets the same level of reader-interest as cute graphics shouting about The League of Shame (Mail, again).

Of course, social reseach shouldn’t (can’t?) be held up to the same level of scrutiny as pure scientific research, simply because of the epistemological debate inherent in answering any social questions, but sometimes a line has to be drawn between valid qualitative assessment and, well, pure unadulterated bollocks.

Some legal nonsense…

Classic first post trite…which is no doubt going to prove ultimately redundant, other than to remind any readers who may find themselves here, that this blog is brought to you in association with the Local Government Act:

“Postholders are not permitted to speak in public where there is an apparent intention to affect public support for a political party.”

So, if anyone spots any apparent intentions, then please. Please. Let me know.