Do crime statistics gloss over people's experiences in high-crime areas? Photograph: Graham Turner
Effective government depends upon representative statistics for efficient policy making. Britain's rulers have known this since before the time of the Domesday Book. Yet Home Office sampling in national crime surveys does not reflect many people's experience of crime.
Home Office annual reports on national crime figures include crime recorded by police and British Crime Survey data for England and Wales. This data does not reveal the true picture of crime that is concentrated in high-crime neighbourhoods - such as some notorious housing estates - nor renowned wealthy and elite neighbourhoods. And this means the concentration of victimisation and self-reported offending between contrasting areas is hidden from the public.
The true picture of the misery of people living in high-crime neighbourhoods is hidden by mixing their particularly concentrated crime experience in with that of people living in nice safe villages and leafy low-crime suburbs.
Examining the data by police beats does not solve this problem, because very often these areas include part of one high-crime estate and part of another totally distinct very low-crime neighbourhood. The result again is most likely to be statistical homogenisation.
Attempting to be representative of all sections of society, the British Crime Survey tries to measure the experiences of people living in different types of housing - but the data is sampled and broken down by commercial market research systems which group people together into socio-economic categories, not by known geographically determined neighbourhoods. That makes the results socially spurious.
Why not have annual surveys of crime rates in our most notorious neighbourhoods? Could it be because the news would be bad and difficult to manage politically?
To get a truer picture of crime requires a "notorious neighbourhoods" and a "wealthy elite neighbourhoods" sampled crime survey, comprising national data that can be stratified on both a local and national basis. Then let us see how people's experience and fear of different crimes, and the self-reported range of offences (including theft, violence and fraud), is going up or down on an annual basis between the most and least desirable places to live.
Fear of stigmatising high crime neighbourhoods should not be taken as an excuse for inaction here. Those living in such areas cannot be stigmatised much more than they already are.
These arguments are published in a book based on recommendations of criminologists attending a private seminar in Windsor last year - sponsored, ironically, by the Home Office: Hough and Maxfield (2007), Surveying Crime in the 21st Century. A chapter by Dr Mike Sutton stridently calls for the Home Office to stop offering the public a diet of weak crime figures.
· The full length version of this blog can be read at Bent Society
