Statistics and Damned Lies

A month ago Chris Grayling, the Tories crime spokesman, was rebuked by Sir Michael Scholar, chair of the UK Statistics Authority for misusing crime figures to attempt to show that violent crime has risen.

Not deterred, Grayling is back today claiming again that violent crime is on the rise, this time with the use of House of Commons Statistics. These, he says show a 44% rise in violent crime since Labour came to power.

The full numbers have not yet been released by the Tories, so we have to take it on trust.

But Sir Michaels Scholar has written to Grayling again. Apparently Grayling has been seeking assurances that if  issues more statistics, Sir Michael will not intervene again. Sir Michael gives no such assurance..... as the letter below shows;

Chair of the UK Statistics Authority, Sir Michael Scholar KCB
Chris Grayling MP
House of Commons
LONDON
SW1A 0AA
8 March 2010

Dear Mr Grayling

VIOLENT CRIME STATISTICS

Thank you for your letter of 5 March.
I am very grateful for your assurance that you are keen to use the available statistics correctly. The UK Statistics Authority, for its part, is keen to help as much as it can.
You asked me to confirm that there will be no further intervention from the Authority if you make public the views expressed in your letter, based, as you record them to be, on advice from the House of Commons Library. I am afraid that it will not be possible for the Authority to give advance clearance to your, or anyone else’s, future statements on this subject – or indeed on any other. In assessing whether statistics are being used properly, and not in a misleading way, much depends on the accuracy of the language which is used, the context, and the inferences which are drawn from the statistics. On some issues – including violent crime – there are several series of statistics which attempt to measure the same phenomenon using different methodologies, and which can sometimes produce results which appear to point in different directions. In such cases, the selective quotation of one without regard to the other could prove misleading, and a balanced presentation of an inevitably complex case would refer to all available statistics, and the uncertainties and ambiguities which they sometimes reveal. The Home Office’s statistical publications document these issues.
This brings to me to the particular case of violent crime statistics. I am glad to learn that you have now sought and been given guidance on these statistics from the House of Commons Library. The adjustment you record them as having suggested to you in relation to the police recorded statistics certainly provides a more accurate comparison than can be made through the unqualified use of published numbers a decade apart, without regard to the definitional changes which have taken place during that time. But a more balanced commentary on national trends in violent crime would, in the view of the Authority, also make reference to the estimates given in the British Crime Survey, which in our view provide a more reliable measure of the national trend over time. All this was set out in the Authority’s Note which I attached to my letter to you of 4 February.
Finally, I am afraid I cannot allow your assertion to go unchallenged that you used no statistics in your BBC radio interview. The transcript available to me (and attached) records you as saying: “If you go into depths of the figures you’ll find a 98% increase in serious violent crime which was almost unaffected by any changes at all”.
I hope you will find my comments helpful as you consider how in future to present the statistics on violent crime. The Authority will not comment on the use of official statistics unless we judge it clearly necessary – as we did when I wrote to you on 4 February – to do so in order to prevent damage to trust in official statistics.
Yours sincerely
Sir Michael Scholar KCB

The key bit that I note is;

a more balanced commentary on national trends in violent crime would, in the view of the Authority, also make reference to the estimates given in the British Crime Survey, which in our view provide a more reliable measure of the national trend over time. All this was set out in the Authority’s Note which I attached to my letter to you of 4 February.

 So Sir Michael thinks the BCS is a more reliable measure of the national trend....

And Chris Grayling, Tory MP, who has already been rebuked for misusing statistics on the same subject, disagrees..

Who would you believe...?