Search This Blog

Wednesday, 8 June 2011

Does the National Statistician understand crime data?

I have broken my silence because I am seriously concerned. The National Statistician's Review of Crime Statistics in England and Wales  was published yesterday. This is what my last post but one referred to. I got the person of the National Statistician wrong. It is now Jil Matheson pictured above with Prime minister David Cameron and on her own.

Unsurprising she has recommended that the Office of National Statistics should take responsibility for crime statistics away from the Home Office. This recommendation 5. Recommendation 1 is:

"The body responsible for the publication of crime statistics should seek to improve the presentation of the statistics to give users and the public a clearer understanding of the overall picture of crime, by providing the major and other sources of crime statistics together with additional contextual information."

This will only happen if the statisticians understand the raw data they present. Figure 1 shown above of the report and the explanatory paragraph, clearly shows they do not understand the scope of police incidents. At best the paragraph and figure present a confusing, unclear explanation. In fact it is totally wrong, worryingly wrong, because whoever wrote paragraph and compiled the figure does not have the first clue about the reporting mechanisms resulting in police recorded crime and the contents of the incident database.

The incident database has details of less than 50% of notifiable offences that end up being counted as police recorded crime but contains hundreds of thousands of incidents that come within recorded crime categories that are not counted as police recorded crime. Additionally not all of the 4.3 million notifiable offences are mapped because either they have no mappable location or the fall outside the categorises that are mapped.