Today 07/07/10. I have just finished listening to Prime Minister’s Question Time.
David Cameron answered a question from Harriet Harman on Crime with a whole string of statistics to show that violent crime had risen under Labour. Harriet Harman, in my view rightly, vigorously contested this.
I am seriously concerned by this.
During the early days of the general election Campaign a row broke out when Chris Grayling, who was then the Shadow home secretary came out with what appears to be very similar assertions.
These were picked up by the reporter Mark Easton who established that the assertions were based on the comparison of two sets of figures which should not have been compared.
• The matter was explored in an interview with Chris Grayling on BBC Today. http://news.bbc.co.uk/today/hi/today/newsid_8494000/8494982.stm
The Chair of the UK Statistics Authority, Sir Michael Scholar was seriously concerned that using statistics in this way could have the effect of undermining trust in crime statistics and published the following:
Doing damage to trust in statistics has serious consequences. If we reach a point where each side is only prepared to trust the statistics it likes it becomes impossible for us to engage in calm pragmatic analysis of what the problems are and what needs to be done.
Effective politics is about being able to talk to people that we do not necessarily like. Good statistics is a tool to make this possible. Bad statistics can have the effect of locking people into polarised positions. The stalemate that has happened in this PMQs and the previous one is a clear example of that.
We all like statistics which confirm our own position! This article explains why. http://is.gd/deeTc persuading people to shift from believing what they want to believe to what is the truth will often rely on people being able to trust the information that is given to them.
This government has a problem. I do not believe what I am being told. And I do not think that I am alone in that!
I need to see the government move to a much more careful and well advised use of statistical material before I can begin to hear what they are trying to say!