Disputing Violent Crime Statistics

David Cameron blasted for using dodgy crime figures to mislead MPs

Daily Mirror 15/7/10

This is a story that has been rumbling in the background for almost a year now. I have not specifically commented on it up to now because I really need to write a paper of about 20,000 words to explain why everyone is right but at the same time everyone has got certain things wrong. What has prompted me to comment now is that the government says it is going to conduct a review of the crime figures.

Where do I start? Probably best at the publication that is in dispute. This can be found here. This is a Home Office publication explaining the findings of the The British Crime Survey (BCS) and police recorded crime. Ironically, given the fact that it is being slammed by the new government, this is by far the best document of this nature that has ever been produced. For the first time the limitations of the statistics are accurately and comprehensively outlined. This is a grown up document for grown up people. The strength and weakness of the document is that it is researched and compiled by statisticians and quantitative social scientists. They are on firm ground when dealing with the BCS but they do not understand the subtleties of crime definitions and police recording practices of police recorded crime. Leading to the presentation of confusing police recorded crime statistics especially those relating to violence.

This is what the BCS shows regarding violent crime trends, clearly showing a increase up to 1997 when the Conservatives were in power and a decrease since 1997 when Labour was in power. Now this looks all too much like Labour Party spin doctor craftiness to the Conservatives. So they looked deep into the document to find police recorded crime statistics that appear to contradict the BCS figures. I have created a simple graph of these statistics that relate to violent crimes with injury.

The document clearly explains that because of different police counting rules this trend is misleading. The real problem is to do with the haphazard categorisation of crimes that mean that many crimes that are violent are not included in the violent crime figures. Some crimes that most people would not consider violent are. And the whole concept of violence with and without injury is often legally inaccurate. These are areas that the researchers and statisticians appear blind of. There are people in the police who know what is going on but as long as crime figures, especially those relating to violence, are performance indicators for the police it is not in their interest to let on.
 
I will explain further in future blogs.