There are "lies, damned lies and crime statistics", as Alan Travis wrote for us. And the latest figures for England and Wales are out today.
The new figures from the Home Office The murder rate in England and Wales rose by 5% last year to 636 but overall crime remained stable or even fell despite the August riots, according to Home Office figures. The murder rise is largely down to the Cumbria shootings which killed 12 people.
Murder statistics often confound received wisdom. The figures this year show:
? Of the 636 murders in 2010/11, the biggest single number used a sharp instrument, such as a knife, for the murder weapon
? The vast majority of child murder victims are killed by a parent - 36 out of 56. In total, 43 of those victims knew their murderer
? The same goes with total murders - 64% were known to their victim, 33% were friends
? 60% of murders are caused by a quarrel and loss of temper. Only 3% are as a result of theft or robbery
The Home Office said the August riots had only a small impact on the overall crime rate, partly because of the way the official counting rules operate. The official police figures link about 5,000 offences to the riots, including 184 incidents of violent disorder, but do not actually class any of them as riot
Other key findings are:
? Police figures show a 19% fall in gun crime offences, down from 7,749 to 6,285 offences
? Total of 5,112 crimes were officially recorded as directly relating to the riots between 6 and 11 August last year compared with a total of 344,937 offences recorded across England and Wales during that month
? This compares with a total of 4,062,866 crimes recorded by the police in the 12 months to September 2011
? The latest BCS figures showed a significant rise in the number of people who say police and councils are doing a good job dealing with crime and anti-social behaviour ? up to 57% of all those questioned, from just over half a year ago
British crime statistics are complicated partly because of double recording. Firstly, there are the official police figures (which historically under-record the true level of crime). Then there is the British Crime Survey - 50,000 people are asked each year for their direct experiences of crime. This is regarded as the most authoritative of either.
The police detection rate ? meaning that a suspect has been identified and interviewed and there is sufficient evidence to bring a charge ? remained at 28% in 2010/11. So, under a third of all crimes get solved.
But one of the most interesting figures is about the perception of crime. The British Crime Survey asks people whether they think crime is getting worse where they live and nationally. So, people think crime is getting worse - but not where they live. It's the perception gap between what we know is going on and what we think is going on. Or is that the Daily Mail gap?
The report published last summer said that:
There remains a disparity between perceived likelihood and actual prevalence of crime. For example, 13% of respondents thought that they were very likely or fairly likely to be a victim of violent crime in the next 12 months, compared with three per cent who reported having been a victim of such a crime in the year before interview