The war of numbers
It's election time. Two unsolicited matey e-mails have dropped into my inbox from opposing political advisers, both wanting to draw my attention to Home Office statistics.
A Labour insider helpfully sends this graph showing how crime is down and fear of crime is rising.
Nothing new in the data with which I am familiar. But the sub-text seems clear enough. "Don't believe all that Tory nonsense about crime going up. They are playing on people's fears".
Meanwhile a Conservative advisor points me in the direction of an official document "found" on the internet which suggests Home Office press releases are full of "statistical omissions".
The sub-text here is that "the Labour government has been playing fast and loose with crime figures and cannot be trusted".
I sense that we are all being "softened up" for the battles to come on crime during the campaign. But it is also a clue as to how important numbers are going to be in the election arguments about criminal justice.
The criticism by the UK Statistics Authority of officials in No 10 over knife crime figures and, more recently, shadow home secretary Chris Grayling on violent crime stats (both following revelations on this blog) has focused election strategists' minds on the power of data.
The Labour prompting around crime and the fear of crime is designed to imply that the government can claim credit for falling levels of victimhood (a controversial assertion) and that high and arguably irrational levels of fear is nothing to do with politicians (equally controversial).
That experience of crime has fallen significantly since 1995 is difficult to contest, but the cause may have little or nothing to do with a thousand government policy initiatives.
Technological and economic factors such as security systems and the fall in the price of consumer durables like DVDs may have had a far greater important impact on volume crimes than the dozens of criminal justice laws introduced.
Equally, it could be argued that much government activity around crime has had the unintended consequence of making people feel more anxious.
The Conservative nudging towards the work of the Home Office's Surveys, Design and Statistics Subcommittee inspired the Daily Mail to assert that "nearly two-thirds of Government press releases contain misleading or unsubstantiated claims".
However, the document actually says that its findings do "not necessarily mean that the statistics were misleading or inaccurate". According to one of the statisticians who produced the "damning report", it is not damning in the least.
Professor Sheila Bird, who will be presenting the results at the Royal Statistical Society tomorrow, tells me that the findings are "not shocking at all but a statistical standard to aim for".
I sometimes feel that elections are the opposite of what they claim to be: far from offering an opportunity for Britain to make a reasoned judgement about the policies to guide our country over the next five years, we witness a crazy scramble for power where truth and perspective, as in a real war, are early victims.
In that context, perhaps, a debate about numbers is a step in the right direction.