The new Government promised a change: but in the way it is handling statistics, it has so far been a change for the worse.
 
The latest embarrassment is a letter from Sir Michael Scholar, chair of the UK Statistics Authority, to Alan Johnson MP, supporting his claim that the Prime Minister misrepresented levels of violent crime and gun crime in the House of Commons on July 7. David Cameron said that violent crime had doubled, a claim not borne out by the British Crime Survey or (as far as comparisons are possible) by police-recorded crime statistics.
 
The latest statistics, out today,  show a 9 per cent fall in all crime as measured by the British Crime Survey and an 8 per cent fall in police-recorded crime. Violent crime shows a non-significant 1 per cent decline in the BCS and police-recorded crime a 4 per cent fall in violence against the person. The BCS is the best source for long-term trends and shows no evidence to justify Cameron’s claims. See graph below:
 
     
 
The same claims were made before the election by Chris Grayling MP, rebutted by the UKSA, then repeated by Mr Grayling on the say-so of a report he allegedly acquired from the House of Commons library which to my knowledge has ever been published. At the time I asked Mr Grayling’s office for a copy but received no reply.
 
Yesterday I blogged about the unemployment estimates released by the Office of  Budget Responsibility in a highly irresponsible fashion, again to be used by Mr Cameron in the House of Commons.
 
Times have changed since 1997. Governments that play fast and loose with statistics risk losing their reputation for honesty, pretty important for this Government if it is to cut public spending without triggering unrest. It’s time ministers worked that out for themselves, but there's no harm reminding them.