Statistical Blunders and Manipulations
David Green, 7 October 2004
Something has gone badly wrong with the Home Office statistical service. Not all that long ago you could place reasonable reliance on official statistics. Today a mixture of bungling and deceit seem to be the rule.
I recently sent an email to the Home Office about the prison statistics asking them to clarify a figure on page 173 (Table 9.10) of Prison Statistics 2002. One row refers to ‘Males aged under 17’ and the next row to ‘Males aged 18-20’. If interpreted literally,this means that the figures for males aged 17 are not included.
I asked the Home Office if the phrase ‘Males aged under 17’ should be ‘Males aged 17 and under’?
They replied a few weeks later by saying that ‘under 17’ means under 17.999 recurring. Here is the exact quotation:
“In response to your recent enquiry regarding the above, the ‘aged under 17’ refers to those aged 17 or under in whole years (so up to 17.999….. years).”
Unless I am sadly mistaken, it would surely be much clearer to say ‘under 18’.
This was probably an example of bungling, but the Home Office is also not above statistical manipulation that can not be put down to error. Take a look at this short report on the Government’s policy targets.